jansson 2.2.1

For maximum compatibility.

.gitignore

@@ -1,3 +1,4 @@
+*~
 *.o
 *.a
 .libs
@@ -23,3 +24,4 @@ missing
 stamp-h1
 *.pyc
 *.pc
+/src/jansson_config.h

CHANGES

@@ -1,3 +1,240 @@
+Version 2.2.1
+=============
+
+Released 2011-10-06
+
+* Bug fixes:
+
+  - Fix real number encoding and decoding under non-C locales. (#32)
+
+  - Fix identifier decoding under non-UTF-8 locales. (#35)
+
+  - `json_load_file()`: Open the input file in binary mode for maximum
+    compatiblity.
+
+* Documentation:
+
+  - Clarify the lifecycle of the result of the ``s`` fromat of
+    `json_unpack()`. (#31)
+
+  - Add some portability info. (#36)
+
+  - Little clarifications here and there.
+
+* Other:
+
+  - Some style fixes, issues detected by static analyzers.
+
+
+Version 2.2
+===========
+
+Released 2011-09-03
+
+* New features:
+
+  - `json_dump_callback()`: Pass the encoder output to a callback
+    function in chunks.
+
+* Bug fixes:
+
+  - `json_string_set()`: Check that target is a string and value is
+    not NULL.
+
+* Other:
+
+  - Documentation typo fixes and clarifications.
+
+
+Version 2.1
+===========
+
+Released 2011-06-10
+
+* New features:
+
+  - `json_loadb()`: Decode a string with a given size, useful if the
+    string is not null terminated.
+
+  - Add ``JSON_ENCODE_ANY`` encoding flag to allow encoding any JSON
+    value. By default, only arrays and objects can be encoded. (#19)
+
+  - Add ``JSON_REJECT_DUPLICATES`` decoding flag to issue a decoding
+    error if any JSON object in the input contins duplicate keys. (#3)
+
+  - Add ``JSON_DISABLE_EOF_CHECK`` decoding flag to stop decoding after a
+    valid JSON input. This allows other data after the JSON data.
+
+* Bug fixes:
+
+  - Fix an additional memory leak when memory allocation fails in
+    `json_object_set()` and friends.
+
+  - Clear errno before calling `strtod()` for better portability. (#27)
+
+* Building:
+
+  - Avoid set-but-not-used warning/error in a test. (#20)
+
+* Other:
+
+  - Minor clarifications to documentation.
+
+
+Version 2.0.1
+=============
+
+Released 2011-03-31
+
+* Bug fixes:
+
+  - Replace a few `malloc()` and `free()` calls with their
+    counterparts that support custom memory management.
+
+  - Fix object key hashing in json_unpack() strict checking mode.
+
+  - Fix the parentheses in ``JANSSON_VERSION_HEX`` macro.
+
+  - Fix `json_object_size()` return value.
+
+  - Fix a few compilation issues.
+
+* Portability:
+
+  - Enhance portability of `va_copy()`.
+
+  - Test framework portability enhancements.
+
+* Documentation:
+
+  - Distribute ``doc/upgrading.rst`` with the source tarball.
+
+  - Build documentation in strict mode in ``make distcheck``.
+
+
+Version 2.0
+===========
+
+Released 2011-02-28
+
+This release is backwards incompatible with the 1.x release series.
+See the chapter "Upgrading from older versions" in documentation for
+details.
+
+* Backwards incompatible changes:
+
+  - Unify unsigned integer usage in the API: All occurences of
+    unsigned int and unsigned long have been replaced with size_t.
+
+  - Change JSON integer's underlying type to the widest signed integer
+    type available, i.e. long long if it's supported, otherwise long.
+    Add a typedef json_int_t that defines the type.
+
+  - Change the maximum indentation depth to 31 spaces in encoder. This
+    frees up bits from the flags parameter of encoding functions
+    `json_dumpf()`, `json_dumps()` and `json_dump_file()`.
+
+  - For future needs, add a flags parameter to all decoding functions
+    `json_loadf()`, `json_loads()` and `json_load_file()`.
+
+* New features
+
+  - `json_pack()`, `json_pack_ex()`, `json_vpack_ex()`: Create JSON
+    values based on a format string.
+
+  - `json_unpack()`, `json_unpack_ex()`, `json_vunpack_ex()`: Simple
+    value extraction and validation functionality based on a format
+    string.
+
+  - Add column, position and source fields to the ``json_error_t``
+    struct.
+
+  - Enhance error reporting in the decoder.
+
+  - ``JANSSON_VERSION`` et al.: Preprocessor constants that define the
+    library version.
+
+  - `json_set_alloc_funcs()`: Set custom memory allocation functions.
+
+* Fix many portability issues, especially on Windows.
+
+* Configuration
+
+  - Add file ``jansson_config.h`` that contains site specific
+    configuration. It's created automatically by the configure script,
+    or can be created by hand if the configure script cannot be used.
+    The file ``jansson_config.h.win32`` can be used without
+    modifications on Windows systems.
+
+  - Add a section to documentation describing how to build Jansson on
+    Windows.
+
+  - Documentation now requires Sphinx 1.0 or newer.
+
+
+Version 1.3
+===========
+
+Released 2010-06-13
+
+* New functions:
+
+  - `json_object_iter_set()`, `json_object_iter_set_new()`: Change
+    object contents while iterating over it.
+
+  - `json_object_iter_at()`: Return an iterator that points to a
+    specific object item.
+
+* New encoding flags:
+
+  - ``JSON_PRESERVE_ORDER``: Preserve the insertion order of object
+    keys.
+
+* Bug fixes:
+
+  - Fix an error that occured when an array or object was first
+    encoded as empty, then populated with some data, and then
+    re-encoded
+
+  - Fix the situation like above, but when the first encoding resulted
+    in an error
+
+* Documentation:
+
+  - Clarify the documentation on reference stealing, providing an
+    example usage pattern
+
+
+Version 1.2.1
+=============
+
+Released 2010-04-03
+
+* Bug fixes:
+
+  - Fix reference counting on ``true``, ``false`` and ``null``
+  - Estimate real number underflows in decoder with 0.0 instead of
+    issuing an error
+
+* Portability:
+
+  - Make ``int32_t`` available on all systems
+  - Support compilers that don't have the ``inline`` keyword
+  - Require Autoconf 2.60 (for ``int32_t``)
+
+* Tests:
+
+  - Print test names correctly when ``VERBOSE=1``
+  - ``test/suites/api``: Fail when a test fails
+  - Enhance tests for iterators
+  - Enhance tests for decoding texts that contain null bytes
+
+* Documentation:
+
+  - Don't remove ``changes.rst`` in ``make clean``
+  - Add a chapter on RFC conformance
+
+
 Version 1.2
 ===========
 

LICENSE

@@ -1,4 +1,4 @@
-Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

Makefile.am

@@ -1,7 +1,10 @@
 EXTRA_DIST = CHANGES LICENSE README.rst
 SUBDIRS = doc src test
 
-check-local: html
+# "make distcheck" builds the dvi target, so use it to check that the
+# documentation is built correctly.
+dvi:
+	$(MAKE) SPHINXOPTS_EXTRA=-W html
 
 pkgconfigdir = $(libdir)/pkgconfig
 pkgconfig_DATA = jansson.pc

README.rst

@@ -24,7 +24,13 @@ Compilation and Installation
 If you obtained a source tarball, just use the standard autotools
 commands::
 
-   $ ./configure && make && make install
+   $ ./configure
+   $ make
+   $ make install
+
+To run the test suite, invoke::
+
+   $ make check
 
 If the source has been checked out from a Git repository, the
 ./configure script has to be generated fist. The easiest way is to use
@@ -32,28 +38,22 @@ autoreconf::
 
    $ autoreconf -i
 
-To run the test suite, invoke::
-
-   $ make check
-
 
 Documentation
 -------------
 
-Documentation is in the ``doc/`` subdirectory. It's written in
-reStructuredText_ with Sphinx_ annotations, so reading it in plain may
-be inconvenient. For this reason, prebuilt HTML documentation is
-available at http://www.digip.org/jansson/doc/.
+Prebuilt HTML documentation is available at
+http://www.digip.org/jansson/doc/.
 
-To generate HTML documentation yourself, invoke::
+The documentation source is in the ``doc/`` subdirectory. To generate
+HTML documentation, invoke::
 
-   make html
+   $ make html
 
-and point your browser to ``doc/_build/html/index.html``. Sphinx_ is
-required to generate the documentation.
+Then, point your browser to ``doc/_build/html/index.html``. Sphinx_
+1.0 or newer is required to generate the documentation.
 
 
 .. _Jansson: http://www.digip.org/jansson/
 .. _`MIT license`: http://www.opensource.org/licenses/mit-license.php
-.. _reStructuredText: http://docutils.sourceforge.net/rst.html
 .. _Sphinx: http://sphinx.pocoo.org/

configure.ac

@@ -1,5 +1,5 @@
-AC_PREREQ([2.59])
-AC_INIT([jansson], [1.2], [petri@digip.org])
+AC_PREREQ([2.60])
+AC_INIT([jansson], [2.2.1], [petri@digip.org])
 
 AM_INIT_AUTOMAKE([1.10 foreign])
 
@@ -9,20 +9,45 @@ AC_CONFIG_HEADERS([config.h])
 # Checks for programs.
 AC_PROG_CC
 AC_PROG_LIBTOOL
+AM_CONDITIONAL([GCC], [test x$GCC = xyes])
 
 # Checks for libraries.
 
 # Checks for header files.
+AC_CHECK_HEADERS([locale.h])
 
 # Checks for typedefs, structures, and compiler characteristics.
+AC_TYPE_INT32_T
+
+AC_TYPE_LONG_LONG_INT
+case $ac_cv_type_long_long_int in
+     yes) json_have_long_long=1;;
+     *) json_have_long_long=0;;
+esac
+AC_SUBST([json_have_long_long])
+
+AC_C_INLINE
+case $ac_cv_c_inline in
+    yes) json_inline=inline;;
+    no) json_inline=;;
+    *) json_inline=$ac_cv_c_inline;;
+esac
+AC_SUBST([json_inline])
 
 # Checks for library functions.
+AC_CHECK_FUNCS([setlocale localeconv])
+case "$ac_cv_header_locale_h$ac_cv_func_localeconv" in
+     yesyes) json_have_localeconv=1;;
+     *) json_have_localeconv=0;;
+esac
+AC_SUBST([json_have_localeconv])
 
 AC_CONFIG_FILES([
         jansson.pc
         Makefile
         doc/Makefile
         src/Makefile
+        src/jansson_config.h
         test/Makefile
         test/bin/Makefile
         test/suites/Makefile

doc/Makefile.am

@@ -1,8 +1,9 @@
-EXTRA_DIST = conf.py apiref.rst changes.rst gettingstarted.rst		\
-	github_commits.c index.rst tutorial.rst ext/refcounting.py
+EXTRA_DIST = conf.py apiref.rst changes.rst conformance.rst		\
+	gettingstarted.rst github_commits.c index.rst portability.rst	\
+	tutorial.rst upgrading.rst ext/refcounting.py
 
 SPHINXBUILD = sphinx-build
-SPHINXOPTS = -d _build/doctrees -W
+SPHINXOPTS = -d _build/doctrees $(SPHINXOPTS_EXTRA)
 
 html-local:
 	$(SPHINXBUILD) -b html $(SPHINXOPTS) $(srcdir) _build/html
@@ -16,4 +17,4 @@ uninstall-local:
 
 clean-local:
 	rm -rf _build
-	rm -f ext/refcounting.pyc changes.rst
+	rm -f ext/refcounting.pyc

doc/conf.py

@@ -1,13 +1,10 @@
 # -*- coding: utf-8 -*-
 #
 # Jansson documentation build configuration file, created by
-# sphinx-quickstart on Thu Jul 30 11:35:32 2009.
+# sphinx-quickstart on Sun Sep  5 21:47:20 2010.
 #
 # This file is execfile()d with the current directory set to its containing dir.
 #
-# The contents of this file are pickled, so don't put values in the namespace
-# that aren't pickleable (module imports are okay, they're removed automatically).
-#
 # Note that not all possible configuration values are present in this
 # autogenerated file.
 #
@@ -15,44 +12,45 @@
 # serve to show the default.
 
 import sys, os
-sys.path.insert(0, os.path.abspath('ext'))
 
-# If your extensions (or modules documented by autodoc) are in another directory,
+# If extensions (or modules to document with autodoc) are in another directory,
 # add these directories to sys.path here. If the directory is relative to the
 # documentation root, use os.path.abspath to make it absolute, like shown here.
-#sys.path.append(os.path.abspath('.'))
+sys.path.insert(0, os.path.abspath('ext'))
 
-# General configuration
-# ---------------------
+# -- General configuration -----------------------------------------------------
+
+# If your documentation needs a minimal Sphinx version, state it here.
+needs_sphinx = '1.0'
 
 # Add any Sphinx extension module names here, as strings. They can be extensions
 # coming with Sphinx (named 'sphinx.ext.*') or your custom ones.
 extensions = ['refcounting']
 
 # Add any paths that contain templates here, relative to this directory.
-templates_path = []
+templates_path = ['_templates']
 
 # The suffix of source filenames.
 source_suffix = '.rst'
 
 # The encoding of source files.
-#source_encoding = 'utf-8'
+#source_encoding = 'utf-8-sig'
 
 # The master toctree document.
 master_doc = 'index'
 
 # General information about the project.
 project = u'Jansson'
-copyright = u'2009, Petri Lehtinen'
+copyright = u'2009-2011, Petri Lehtinen'
 
 # The version info for the project you're documenting, acts as replacement for
 # |version| and |release|, also used in various other places throughout the
 # built documents.
 #
 # The short X.Y version.
-version = '1.2'
+version = '2.2.1'
 # The full version, including alpha/beta/rc tags.
-release = '1.2'
+release = version
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.
@@ -64,15 +62,13 @@ release = '1.2'
 # Else, today_fmt is used as the format for a strftime call.
 #today_fmt = '%B %d, %Y'
 
-# List of documents that shouldn't be included in the build.
-#unused_docs = []
-
-# List of directories, relative to source directory, that shouldn't be searched
-# for source files.
-exclude_trees = ['_build']
+# List of patterns, relative to source directory, that match files and
+# directories to ignore when looking for source files.
+exclude_patterns = ['_build']
 
 # The reST default role (used for this markup: `text`) to use for all documents.
-default_role = 'cfunc'
+default_role = 'c:func'
+primary_domain = 'c'
 
 # If true, '()' will be appended to :func: etc. cross-reference text.
 #add_function_parentheses = True
@@ -88,14 +84,23 @@ default_role = 'cfunc'
 # The name of the Pygments (syntax highlighting) style to use.
 pygments_style = 'sphinx'
 
+# A list of ignored prefixes for module index sorting.
+#modindex_common_prefix = []
 
-# Options for HTML output
-# -----------------------
 
-# The style sheet to use for HTML and HTML Help pages. A file of that name
-# must exist either in Sphinx' static/ path, or in one of the custom paths
-# given in html_static_path.
-html_style = 'default.css'
+# -- Options for HTML output ---------------------------------------------------
+
+# The theme to use for HTML and HTML Help pages.  See the documentation for
+# a list of builtin themes.
+#html_theme = 'default'
+
+# Theme options are theme-specific and customize the look and feel of a theme
+# further.  For a list of options available for each theme, see the
+# documentation.
+#html_theme_options = {}
+
+# Add any paths that contain custom themes here, relative to this directory.
+#html_theme_path = []
 
 # The name for this set of Sphinx documents.  If None, it defaults to
 # "<project> v<release> documentation".
@@ -116,7 +121,7 @@ html_style = 'default.css'
 # Add any paths that contain custom static files (such as style sheets) here,
 # relative to this directory. They are copied after the builtin static files,
 # so a file named "default.css" will overwrite the builtin "default.css".
-html_static_path = []
+#html_static_path = ['_static']
 
 # If not '', a 'Last updated on:' timestamp is inserted at every page bottom,
 # using the given strftime format.
@@ -134,7 +139,7 @@ html_static_path = []
 #html_additional_pages = {}
 
 # If false, no module index is generated.
-#html_use_modindex = True
+#html_domain_indices = True
 
 # If false, no index is generated.
 #html_use_index = True
@@ -142,23 +147,28 @@ html_static_path = []
 # If true, the index is split into individual pages for each letter.
 #html_split_index = False
 
-# If true, the reST sources are included in the HTML build as _sources/<name>.
-#html_copy_source = True
+# If true, links to the reST sources are added to the pages.
+#html_show_sourcelink = True
+
+# If true, "Created using Sphinx" is shown in the HTML footer. Default is True.
+#html_show_sphinx = True
+
+# If true, "(C) Copyright ..." is shown in the HTML footer. Default is True.
+#html_show_copyright = True
 
 # If true, an OpenSearch description file will be output, and all pages will
 # contain a <link> tag referring to it.  The value of this option must be the
 # base URL from which the finished HTML is served.
 #html_use_opensearch = ''
 
-# If nonempty, this is the file name suffix for HTML files (e.g. ".xhtml").
-#html_file_suffix = ''
+# This is the file name suffix for HTML files (e.g. ".xhtml").
+#html_file_suffix = None
 
 # Output file base name for HTML help builder.
 htmlhelp_basename = 'Janssondoc'
 
 
-# Options for LaTeX output
-# ------------------------
+# -- Options for LaTeX output --------------------------------------------------
 
 # The paper size ('letter' or 'a4').
 #latex_paper_size = 'letter'
@@ -167,10 +177,10 @@ htmlhelp_basename = 'Janssondoc'
 #latex_font_size = '10pt'
 
 # Grouping the document tree into LaTeX files. List of tuples
-# (source start file, target name, title, author, document class [howto/manual]).
+# (source start file, target name, title, author, documentclass [howto/manual]).
 latex_documents = [
-  ('index', 'Jansson.tex', ur'Jansson Documentation',
-   ur'Petri Lehtinen', 'manual'),
+  ('index', 'Jansson.tex', u'Jansson Documentation',
+   u'Petri Lehtinen', 'manual'),
 ]
 
 # The name of an image file (relative to this directory) to place at the top of
@@ -181,6 +191,12 @@ latex_documents = [
 # not chapters.
 #latex_use_parts = False
 
+# If true, show page references after internal links.
+#latex_show_pagerefs = False
+
+# If true, show URL addresses after external links.
+#latex_show_urls = False
+
 # Additional stuff for the LaTeX preamble.
 #latex_preamble = ''
 
@@ -188,4 +204,14 @@ latex_documents = [
 #latex_appendices = []
 
 # If false, no module index is generated.
-#latex_use_modindex = True
+#latex_domain_indices = True
+
+
+# -- Options for manual page output --------------------------------------------
+
+# One entry per manual page. List of tuples
+# (source start file, name, description, authors, manual section).
+man_pages = [
+    ('index', 'jansson', u'Jansson Documentation',
+     [u'Petri Lehtinen'], 1)
+]

doc/conformance.rst

@@ -0,0 +1,111 @@
+.. _rfc-conformance:
+
+***************
+RFC Conformance
+***************
+
+JSON is specified in :rfc:`4627`, *"The application/json Media Type
+for JavaScript Object Notation (JSON)"*.
+
+Character Encoding
+==================
+
+Jansson only supports UTF-8 encoded JSON texts. It does not support or
+auto-detect any of the other encodings mentioned in the RFC, namely
+UTF-16LE, UTF-16BE, UTF-32LE or UTF-32BE. Pure ASCII is supported, as
+it's a subset of UTF-8.
+
+Strings
+=======
+
+JSON strings are mapped to C-style null-terminated character arrays,
+and UTF-8 encoding is used internally. Strings may not contain
+embedded null characters, not even escaped ones.
+
+For example, trying to decode the following JSON text leads to a parse
+error::
+
+    ["this string contains the null character: \u0000"]
+
+All other Unicode codepoints U+0001 through U+10FFFF are allowed.
+
+Unicode normalization or any other transformation is never performed
+on any strings (string values or object keys). When checking for
+equivalence of strings or object keys, the comparison is performed
+byte by byte between the original UTF-8 representations of the
+strings.
+
+Numbers
+=======
+
+Real vs. Integer
+----------------
+
+JSON makes no distinction between real and integer numbers; Jansson
+does. Real numbers are mapped to the ``double`` type and integers to
+the ``json_int_t`` type, which is a typedef of ``long long`` or
+``long``, depending on whether ``long long`` is supported by your
+compiler or not.
+
+A JSON number is considered to be a real number if its lexical
+representation includes one of ``e``, ``E``, or ``.``; regardless if
+its actual numeric value is a true integer (e.g., all of ``1E6``,
+``3.0``, ``400E-2``, and ``3.14E3`` are mathematical integers, but
+will be treated as real values).
+
+All other JSON numbers are considered integers.
+
+When encoding to JSON, real values are always represented
+with a fractional part; e.g., the ``double`` value 3.0 will be
+represented in JSON as ``3.0``, not ``3``.
+
+Overflow, Underflow & Precision
+-------------------------------
+
+Real numbers whose absolute values are too small to be represented in
+a C ``double`` will be silently estimated with 0.0. Thus, depending on
+platform, JSON numbers very close to zero such as 1E-999 may result in
+0.0.
+
+Real numbers whose absolute values are too large to be represented in
+a C ``double`` will result in an overflow error (a JSON decoding
+error). Thus, depending on platform, JSON numbers like 1E+999 or
+-1E+999 may result in a parsing error.
+
+Likewise, integer numbers whose absolute values are too large to be
+represented in the ``json_int_t`` type (see above) will result in an
+overflow error (a JSON decoding error). Thus, depending on platform,
+JSON numbers like 1000000000000000 may result in parsing error.
+
+Parsing JSON real numbers may result in a loss of precision. As long
+as overflow does not occur (i.e. a total loss of precision), the
+rounded approximate value is silently used. Thus the JSON number
+1.000000000000000005 may, depending on platform, result in the
+``double`` value 1.0.
+
+Signed zeros
+------------
+
+JSON makes no statement about what a number means; however Javascript
+(ECMAscript) does state that +0.0 and -0.0 must be treated as being
+distinct values, i.e. -0.0 |not-equal| 0.0. Jansson relies on the
+underlying floating point library in the C environment in which it is
+compiled. Therefore it is platform-dependent whether 0.0 and -0.0 will
+be distinct values. Most platforms that use the IEEE 754
+floating-point standard will support signed zeros.
+
+Note that this only applies to floating-point; neither JSON, C, or
+IEEE support the concept of signed integer zeros.
+
+.. |not-equal| unicode:: U+2260
+
+Types
+-----
+
+No support is provided in Jansson for any C numeric types other than
+``json_int_t`` and ``double``. This excludes things such as unsigned
+types, ``long double``, etc. Obviously, shorter types like ``short``,
+``int``, ``long`` (if ``json_int_t`` is ``long long``) and ``float``
+are implicitly handled via the ordinary C type coercion rules (subject
+to overflow semantics). Also, no support or hooks are provided for any
+supplemental "bignum" type add-on packages.

doc/ext/refcounting.py

@@ -19,7 +19,7 @@
 
        <description of the json_object function>
 
-    :copyright: Copyright 2009 Petri Lehtinen <petri@digip.org>
+    :copyright: Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
     :license: MIT, see LICENSE for details.
 """
 

doc/gettingstarted.rst

@@ -10,6 +10,9 @@ Compiling and Installing Jansson
 The Jansson source is available at
 http://www.digip.org/jansson/releases/.
 
+Unix-like systems
+-----------------
+
 Unpack the source tarball and change to the source directory:
 
 .. parsed-literal::
@@ -37,7 +40,7 @@ please report them.
 
 If you obtained the source from a Git repository (or any other source
 control system), there's no ``./configure`` script as it's not kept in
-version control. To create the script, Autotools needs to be
+version control. To create the script, the build system needs to be
 bootstrapped. There are many ways to do this, but the easiest one is
 to use ``autoreconf``::
 
@@ -51,18 +54,30 @@ used as described above.
 .. _libtool: http://www.gnu.org/software/libtool/
 
 
-Installing Prebuilt Binary Packages
------------------------------------
+Other Systems
+-------------
 
-Binary ``.deb`` packages for Ubuntu are available in `this PPA`_ at
-Launchpad_. Follow the instructions in the PPA ("Technical details
-about this PPA" link) to take the PPA into use. Then install the -dev
-package::
+On Windows and other non Unix-like systems, you may be unable to run
+the ``./configure`` script. In this case, follow these steps. All the
+files mentioned can be found in the ``src/`` directory.
 
-  sudo apt-get install libjansson-dev
+1. Create ``jansson_config.h``. This file has some platform-specific
+   parameters that are normally filled in by the ``./configure``
+   script:
 
-.. _this PPA: http://launchpad.net/~petri/+archive/ppa
-.. _Launchpad: http://launchpad.net/
+   - On Windows, rename ``jansson_config.h.win32`` to ``jansson_config.h``.
+
+   - On other systems, edit ``jansson_config.h.in``, replacing all
+     ``@variable@`` placeholders, and rename the file to
+     ``jansson_config.h``.
+
+2. Make ``jansson.h`` and ``jansson_config.h`` available to the
+   compiler, so that they can be found when compiling programs that
+   use Jansson.
+
+3. Compile all the ``.c`` files (in the ``src/`` directory) into a
+   library file. Make the library available to the compiler, as in
+   step 2.
 
 
 Building the Documentation
@@ -77,14 +92,14 @@ documentation, invoke::
 
    make html
 
-and point your browser to ``doc/_build/html/index.html``. Sphinx_ is
-required to generate the documentation.
+and point your browser to ``doc/_build/html/index.html``. Sphinx_ 1.0
+or newer is required to generate the documentation.
 
 .. _reStructuredText: http://docutils.sourceforge.net/rst.html
 .. _Sphinx: http://sphinx.pocoo.org/
 
 
-Compiling Programs That Use Jansson
+Compiling Programs that Use Jansson
 ===================================
 
 Jansson involves one C header file, :file:`jansson.h`, so it's enough

doc/github_commits.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -96,7 +96,7 @@ static char *request(const char *url)
 
 int main(int argc, char *argv[])
 {
-    unsigned int i;
+    size_t i;
     char *text;
     char url[URL_SIZE];
 
@@ -117,7 +117,7 @@ int main(int argc, char *argv[])
     if(!text)
         return 1;
 
-    root = json_loads(text, &error);
+    root = json_loads(text, 0, &error);
     free(text);
 
     if(!root)

doc/index.rst

@@ -22,6 +22,11 @@ data. Its main features and design principles are:
 Jansson is licensed under the `MIT license`_; see LICENSE in the
 source distribution for details.
 
+Jansson is used in production and its API is stable. It works on
+numerous platforms, including numerous Unix like systems and Windows.
+It's suitable for use on any system, including desktop, server, and
+small embedded systems.
+
 
 .. _`MIT license`: http://www.opensource.org/licenses/mit-license.php
 .. _Jansson: http://www.digip.org/jansson/
@@ -33,7 +38,10 @@ Contents
    :maxdepth: 2
 
    gettingstarted
+   upgrading
    tutorial
+   conformance
+   portability
    apiref
    changes
 

doc/portability.rst

@@ -0,0 +1,42 @@
+***********
+Portability
+***********
+
+Thread safety
+-------------
+
+Jansson is thread safe and has no mutable global state. The only
+exception are the memory allocation functions, that should be set at
+most once, and only on program startup. See
+:ref:`apiref-custom-memory-allocation`.
+
+There's no locking performed inside Jansson's code, so a multithreaded
+program must perform its own locking if JSON values are shared by
+multiple threads. Jansson's reference counting semantics may make this
+a bit harder than it seems, as it's possible to have a reference to a
+value that's also stored inside a list or object. Modifying the
+container (adding or removing values) may trigger concurrent access to
+such values, as containers manage the reference count of their
+contained values. Bugs involving concurrent incrementing or
+decrementing of deference counts may be hard to track.
+
+Locale
+------
+
+Jansson works fine under any locale.
+
+However, if the host program is multithreaded and uses ``setlocale()``
+to switch the locale in one thread while Jansson is currently encoding
+or decoding JSON data in another thread, the result may be wrong or
+the program may even crash.
+
+Jansson uses locale specific functions for certain string conversions
+in the encoder and decoder, and then converts the locale specific
+values to/from the JSON representation. This fails if the locale
+changes between the string conversion and the locale-to-JSON
+conversion. This can only happen in multithreaded programs that use
+``setlocale()``, because ``setlocale()`` switches the locale for all
+running threads, not only the thread that calls ``setlocale()``.
+
+If your program uses ``setlocale()`` as described above, consider
+using the thread-safe ``uselocale()`` instead.

doc/tutorial.rst

@@ -118,7 +118,7 @@ first newline in the commit message::
 The main function follows. In the beginning, we first declare a bunch
 of variables and check the command line parameters::
 
-    unsigned int i;
+    size_t i;
     char *text;
     char url[URL_SIZE];
 
@@ -149,10 +149,10 @@ If an error occurs, our function ``request`` prints the error and
 returns *NULL*, so it's enough to just return 1 from the main
 function.
 
-Next we'll call :cfunc:`json_loads()` to decode the JSON text we got
+Next we'll call :func:`json_loads()` to decode the JSON text we got
 as a response::
 
-    root = json_loads(text, &error);
+    root = json_loads(text, 0, &error);
     free(text);
 
     if(!root)
@@ -162,8 +162,8 @@ as a response::
     }
 
 We don't need the JSON text anymore, so we can free the ``text``
-variable right after decoding it. If :cfunc:`json_loads()` fails, it
-returns *NULL* and sets error information to the :ctype:`json_error_t`
+variable right after decoding it. If :func:`json_loads()` fails, it
+returns *NULL* and sets error information to the :type:`json_error_t`
 structure given as the second parameter. In this case, our program
 prints the error information out and returns 1 from the main function.
 
@@ -182,8 +182,8 @@ First, we'll extract the ``commits`` array from the JSON response::
 
 This is the array that contains objects describing latest commits in
 the repository. We check that the returned value really is an array.
-If the key ``commits`` doesn't exist, :cfunc:`json_object_get()`
-returns *NULL*, but :cfunc:`json_is_array()` handles this case, too.
+If the key ``commits`` doesn't exist, :func:`json_object_get()`
+returns *NULL*, but :func:`json_is_array()` handles this case, too.
 
 Then we proceed to loop over all the commits in the array::
 
@@ -200,9 +200,9 @@ Then we proceed to loop over all the commits in the array::
         }
     ...
 
-The function :cfunc:`json_array_size()` returns the size of a JSON
+The function :func:`json_array_size()` returns the size of a JSON
 array. First, we again declare some variables and then extract the
-i'th element of the ``commits`` array using :cfunc:`json_array_get()`.
+i'th element of the ``commits`` array using :func:`json_array_get()`.
 We also check that the resulting value is a JSON object.
 
 Next we'll extract the commit ID and commit message, and check that
@@ -225,7 +225,7 @@ they both are JSON strings::
 
 And finally, we'll print the first 8 characters of the commit ID and
 the first line of the commit message. A C-style string is extracted
-from a JSON string using :cfunc:`json_string_value()`::
+from a JSON string using :func:`json_string_value()`::
 
         message_text = json_string_value(message);
         printf("%.8s %.*s\n",
@@ -235,9 +235,9 @@ from a JSON string using :cfunc:`json_string_value()`::
     }
 
 After sending the HTTP request, we decoded the JSON text using
-:cfunc:`json_loads()`, remember? It returns a *new reference* to the
+:func:`json_loads()`, remember? It returns a *new reference* to the
 JSON value it decodes. When we're finished with the value, we'll need
-to decrease the reference count using :cfunc:`json_decref()`. This way
+to decrease the reference count using :func:`json_decref()`. This way
 Jansson can release the resources::
 
     json_decref(root);

doc/upgrading.rst

@@ -0,0 +1,76 @@
+.. highlight:: c
+
+******************
+Upgrading from 1.x
+******************
+
+This chapter lists the backwards incompatible changes introduced in
+Jansson 2.0, and the steps that are needed for upgrading your code.
+
+**The incompatibilities are not dramatic.** The biggest change is that
+all decoding functions now require and extra parameter. Most programs
+can be modified to work with 2.0 by adding a ``0`` as the second
+parameter to all calls of :func:`json_loads()`, :func:`json_loadf()`
+and :func:`json_load_file()`.
+
+
+Compatibility
+=============
+
+Jansson 2.0 is backwards incompatible with the Jansson 1.x releases.
+It is ABI incompatible, i.e. all programs dynamically linking to the
+Jansson library need to be recompiled. It's also API incompatible,
+i.e. the source code of programs using Jansson 1.x may need
+modifications to make them compile against Jansson 2.0.
+
+All the 2.x releases are guaranteed to be backwards compatible for
+both ABI and API, so no recompilation or source changes are needed
+when upgrading from 2.x to 2.y.
+
+
+List of Incompatible Changes
+============================
+
+**Decoding flags**
+    For future needs, a ``flags`` parameter was added as the second
+    parameter to all decoding functions, i.e. :func:`json_loads()`,
+    :func:`json_loadf()` and :func:`json_load_file()`. All calls to
+    these functions need to be changed by adding a ``0`` as the second
+    argument. For example::
+
+        /* old code */
+        json_loads(input, &error);
+
+        /* new code */
+        json_loads(input, 0, &error);
+
+
+**Underlying type of JSON integers**
+    The underlying C type of JSON integers has been changed from
+    :type:`int` to the widest available signed integer type, i.e.
+    :type:`long long` or :type:`long`, depending on whether
+    :type:`long long` is supported on your system or not. This makes
+    the whole 64-bit integer range available on most modern systems.
+
+    ``jansson.h`` has a typedef :type:`json_int_t` to the underlying
+    integer type. :type:`int` should still be used in most cases when
+    dealing with smallish JSON integers, as the compiler handles
+    implicit type coercion. Only when the full 64-bit range is needed,
+    :type:`json_int_t` should be explicitly used.
+
+
+**Maximum encoder indentation depth**
+    The maximum argument of the ``JSON_INDENT()`` macro has been
+    changed from 255 to 31, to free up bits from the ``flags``
+    parameter of :func:`json_dumps()`, :func:`json_dumpf()` and
+    :func:`json_dump_file()`. If your code uses a bigger indentation
+    than 31, it needs to be changed.
+
+
+**Unsigned integers in API functions**
+    Version 2.0 unifies unsigned integer usage in the API. All uses of
+    :type:`unsigned int` and :type:`unsigned long` have been replaced
+    with :type:`size_t`. This includes flags, container sizes, etc.
+    This should not require source code changes, as both
+    :type:`unsigned int` and :type:`unsigned long` are usually
+    compatible with :type:`size_t`.

src/Makefile.am

@@ -1,20 +1,26 @@
-include_HEADERS = jansson.h
+include_HEADERS = jansson.h jansson_config.h
 
 lib_LTLIBRARIES = libjansson.la
 libjansson_la_SOURCES = \
 	dump.c \
+	error.c \
 	hashtable.c \
 	hashtable.h \
 	jansson_private.h \
 	load.c \
+	memory.c \
+	pack_unpack.c \
 	strbuffer.c \
 	strbuffer.h \
+	strconv.c \
 	utf.c \
 	utf.h \
-	util.h \
 	value.c
 libjansson_la_LDFLAGS = \
 	-export-symbols-regex '^json_' \
-	-version-info 2:0:2
+	-version-info 6:1:2
 
-AM_CFLAGS = -Wall -Wextra -Werror
+if GCC
+# These flags are gcc specific
+AM_CFLAGS = -Wall -Wextra -Wdeclaration-after-statement -Werror
+endif

src/dump.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -9,7 +9,6 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#include <stdint.h>
 #include <assert.h>
 
 #include <jansson.h>
@@ -20,8 +19,6 @@
 #define MAX_INTEGER_STR_LENGTH  100
 #define MAX_REAL_STR_LENGTH     100
 
-typedef int (*dump_func)(const char *buffer, int size, void *data);
-
 struct string
 {
     char *buffer;
@@ -29,12 +26,12 @@ struct string
     int size;
 };
 
-static int dump_to_strbuffer(const char *buffer, int size, void *data)
+static int dump_to_strbuffer(const char *buffer, size_t size, void *data)
 {
     return strbuffer_append_bytes((strbuffer_t *)data, buffer, size);
 }
 
-static int dump_to_file(const char *buffer, int size, void *data)
+static int dump_to_file(const char *buffer, size_t size, void *data)
 {
     FILE *dest = (FILE *)data;
     if(fwrite(buffer, size, 1, dest) != 1)
@@ -42,10 +39,10 @@ static int dump_to_file(const char *buffer, int size, void *data)
     return 0;
 }
 
-/* 256 spaces (the maximum indentation size) */
-static char whitespace[] = "                                                                                                                                                                                                                                                                ";
+/* 32 spaces (the maximum indentation size) */
+static char whitespace[] = "                                ";
 
-static int dump_indent(unsigned long flags, int depth, int space, dump_func dump, void *data)
+static int dump_indent(size_t flags, int depth, int space, json_dump_callback_t dump, void *data)
 {
     if(JSON_INDENT(flags) > 0)
     {
@@ -67,7 +64,7 @@ static int dump_indent(unsigned long flags, int depth, int space, dump_func dump
     return 0;
 }
 
-static int dump_string(const char *str, int ascii, dump_func dump, void *data)
+static int dump_string(const char *str, int ascii, json_dump_callback_t dump, void *data)
 {
     const char *pos, *end;
     int32_t codepoint;
@@ -154,13 +151,20 @@ static int dump_string(const char *str, int ascii, dump_func dump, void *data)
     return dump("\"", 1, data);
 }
 
-static int object_key_cmp(const void *key1, const void *key2)
+static int object_key_compare_keys(const void *key1, const void *key2)
 {
-    return strcmp(*(const char **)key1, *(const char **)key2);
+    return strcmp((*(const object_key_t **)key1)->key,
+                  (*(const object_key_t **)key2)->key);
 }
 
-static int do_dump(const json_t *json, unsigned long flags, int depth,
-                   dump_func dump, void *data)
+static int object_key_compare_serials(const void *key1, const void *key2)
+{
+    return (*(const object_key_t **)key1)->serial -
+           (*(const object_key_t **)key2)->serial;
+}
+
+static int do_dump(const json_t *json, size_t flags, int depth,
+                   json_dump_callback_t dump, void *data)
 {
     int ascii = flags & JSON_ENSURE_ASCII ? 1 : 0;
 
@@ -179,7 +183,9 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
             char buffer[MAX_INTEGER_STR_LENGTH];
             int size;
 
-            size = snprintf(buffer, MAX_INTEGER_STR_LENGTH, "%d", json_integer_value(json));
+            size = snprintf(buffer, MAX_INTEGER_STR_LENGTH,
+                            "%" JSON_INTEGER_FORMAT,
+                            json_integer_value(json));
             if(size >= MAX_INTEGER_STR_LENGTH)
                 return -1;
 
@@ -190,26 +196,12 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
         {
             char buffer[MAX_REAL_STR_LENGTH];
             int size;
+            double value = json_real_value(json);
 
-            size = snprintf(buffer, MAX_REAL_STR_LENGTH, "%.17g",
-                            json_real_value(json));
-            if(size >= MAX_REAL_STR_LENGTH)
+            size = jsonp_dtostr(buffer, MAX_REAL_STR_LENGTH, value);
+            if(size < 0)
                 return -1;
 
-            /* Make sure there's a dot or 'e' in the output. Otherwise
-               a real is converted to an integer when decoding */
-            if(strchr(buffer, '.') == NULL &&
-               strchr(buffer, 'e') == NULL)
-            {
-                if(size + 2 >= MAX_REAL_STR_LENGTH) {
-                    /* No space to append ".0" */
-                    return -1;
-                }
-                buffer[size] = '.';
-                buffer[size + 1] = '0';
-                size += 2;
-            }
-
             return dump(buffer, size, data);
         }
 
@@ -225,38 +217,44 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
             /* detect circular references */
             array = json_to_array(json);
             if(array->visited)
-                return -1;
+                goto array_error;
             array->visited = 1;
 
             n = json_array_size(json);
 
             if(dump("[", 1, data))
-                return -1;
-            if(n == 0)
+                goto array_error;
+            if(n == 0) {
+                array->visited = 0;
                 return dump("]", 1, data);
+            }
             if(dump_indent(flags, depth + 1, 0, dump, data))
-                return -1;
+                goto array_error;
 
             for(i = 0; i < n; ++i) {
                 if(do_dump(json_array_get(json, i), flags, depth + 1,
                            dump, data))
-                    return -1;
+                    goto array_error;
 
                 if(i < n - 1)
                 {
                     if(dump(",", 1, data) ||
                        dump_indent(flags, depth + 1, 1, dump, data))
-                        return -1;
+                        goto array_error;
                 }
                 else
                 {
                     if(dump_indent(flags, depth, 0, dump, data))
-                        return -1;
+                        goto array_error;
                 }
             }
 
             array->visited = 0;
             return dump("]", 1, data);
+
+        array_error:
+            array->visited = 0;
+            return -1;
         }
 
         case JSON_OBJECT:
@@ -278,48 +276,53 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
             /* detect circular references */
             object = json_to_object(json);
             if(object->visited)
-                return -1;
+                goto object_error;
             object->visited = 1;
 
             iter = json_object_iter((json_t *)json);
 
             if(dump("{", 1, data))
-                return -1;
-            if(!iter)
+                goto object_error;
+            if(!iter) {
+                object->visited = 0;
                 return dump("}", 1, data);
+            }
             if(dump_indent(flags, depth + 1, 0, dump, data))
-                return -1;
+                goto object_error;
 
-            if(flags & JSON_SORT_KEYS)
+            if(flags & JSON_SORT_KEYS || flags & JSON_PRESERVE_ORDER)
             {
-                /* Sort keys */
-
-                const char **keys;
-                unsigned int size;
-                unsigned int i;
+                const object_key_t **keys;
+                size_t size, i;
+                int (*cmp_func)(const void *, const void *);
 
                 size = json_object_size(json);
-                keys = malloc(size * sizeof(const char *));
+                keys = jsonp_malloc(size * sizeof(object_key_t *));
                 if(!keys)
-                    return -1;
+                    goto object_error;
 
                 i = 0;
                 while(iter)
                 {
-                    keys[i] = json_object_iter_key(iter);
+                    keys[i] = jsonp_object_iter_fullkey(iter);
                     iter = json_object_iter_next((json_t *)json, iter);
                     i++;
                 }
                 assert(i == size);
 
-                qsort(keys, size, sizeof(const char *), object_key_cmp);
+                if(flags & JSON_SORT_KEYS)
+                    cmp_func = object_key_compare_keys;
+                else
+                    cmp_func = object_key_compare_serials;
+
+                qsort(keys, size, sizeof(object_key_t *), cmp_func);
 
                 for(i = 0; i < size; i++)
                 {
                     const char *key;
                     json_t *value;
 
-                    key = keys[i];
+                    key = keys[i]->key;
                     value = json_object_get(json, key);
                     assert(value);
 
@@ -327,8 +330,8 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
                     if(dump(separator, separator_length, data) ||
                        do_dump(value, flags, depth + 1, dump, data))
                     {
-                        free(keys);
-                        return -1;
+                        jsonp_free(keys);
+                        goto object_error;
                     }
 
                     if(i < size - 1)
@@ -336,21 +339,21 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
                         if(dump(",", 1, data) ||
                            dump_indent(flags, depth + 1, 1, dump, data))
                         {
-                            free(keys);
-                            return -1;
+                            jsonp_free(keys);
+                            goto object_error;
                         }
                     }
                     else
                     {
                         if(dump_indent(flags, depth, 0, dump, data))
                         {
-                            free(keys);
-                            return -1;
+                            jsonp_free(keys);
+                            goto object_error;
                         }
                     }
                 }
 
-                free(keys);
+                jsonp_free(keys);
             }
             else
             {
@@ -364,18 +367,18 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
                     if(dump(separator, separator_length, data) ||
                        do_dump(json_object_iter_value(iter), flags, depth + 1,
                                dump, data))
-                        return -1;
+                        goto object_error;
 
                     if(next)
                     {
                         if(dump(",", 1, data) ||
                            dump_indent(flags, depth + 1, 1, dump, data))
-                            return -1;
+                            goto object_error;
                     }
                     else
                     {
                         if(dump_indent(flags, depth, 0, dump, data))
-                            return -1;
+                            goto object_error;
                     }
 
                     iter = next;
@@ -384,6 +387,10 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
 
             object->visited = 0;
             return dump("}", 1, data);
+
+        object_error:
+            object->visited = 0;
+            return -1;
         }
 
         default:
@@ -392,38 +399,29 @@ static int do_dump(const json_t *json, unsigned long flags, int depth,
     }
 }
 
-
-char *json_dumps(const json_t *json, unsigned long flags)
+char *json_dumps(const json_t *json, size_t flags)
 {
     strbuffer_t strbuff;
     char *result;
 
-    if(!json_is_array(json) && !json_is_object(json))
-        return NULL;
-
     if(strbuffer_init(&strbuff))
         return NULL;
 
-    if(do_dump(json, flags, 0, dump_to_strbuffer, (void *)&strbuff)) {
-        strbuffer_close(&strbuff);
-        return NULL;
-    }
+    if(json_dump_callback(json, dump_to_strbuffer, (void *)&strbuff, flags))
+        result = NULL;
+    else
+        result = jsonp_strdup(strbuffer_value(&strbuff));
 
-    result = strdup(strbuffer_value(&strbuff));
     strbuffer_close(&strbuff);
-
     return result;
 }
 
-int json_dumpf(const json_t *json, FILE *output, unsigned long flags)
+int json_dumpf(const json_t *json, FILE *output, size_t flags)
 {
-    if(!json_is_array(json) && !json_is_object(json))
-        return -1;
-
-    return do_dump(json, flags, 0, dump_to_file, (void *)output);
+    return json_dump_callback(json, dump_to_file, (void *)output, flags);
 }
 
-int json_dump_file(const json_t *json, const char *path, unsigned long flags)
+int json_dump_file(const json_t *json, const char *path, size_t flags)
 {
     int result;
 
@@ -436,3 +434,13 @@ int json_dump_file(const json_t *json, const char *path, unsigned long flags)
     fclose(output);
     return result;
 }
+
+int json_dump_callback(const json_t *json, json_dump_callback_t callback, void *data, size_t flags)
+{
+    if(!(flags & JSON_ENCODE_ANY)) {
+        if(!json_is_array(json) && !json_is_object(json))
+           return -1;
+    }
+
+    return do_dump(json, flags, 0, callback, data);
+}

src/error.c

@@ -0,0 +1,62 @@
+#include <string.h>
+#include "jansson_private.h"
+
+void jsonp_error_init(json_error_t *error, const char *source)
+{
+    if(error)
+    {
+        error->text[0] = '\0';
+        error->line = -1;
+        error->column = -1;
+        error->position = 0;
+        if(source)
+            jsonp_error_set_source(error, source);
+        else
+            error->source[0] = '\0';
+    }
+}
+
+void jsonp_error_set_source(json_error_t *error, const char *source)
+{
+    size_t length;
+
+    if(!error || !source)
+        return;
+
+    length = strlen(source);
+    if(length < JSON_ERROR_SOURCE_LENGTH)
+        strcpy(error->source, source);
+    else {
+        size_t extra = length - JSON_ERROR_SOURCE_LENGTH + 4;
+        strcpy(error->source, "...");
+        strcpy(error->source + 3, source + extra);
+    }
+}
+
+void jsonp_error_set(json_error_t *error, int line, int column,
+                     size_t position, const char *msg, ...)
+{
+    va_list ap;
+
+    va_start(ap, msg);
+    jsonp_error_vset(error, line, column, position, msg, ap);
+    va_end(ap);
+}
+
+void jsonp_error_vset(json_error_t *error, int line, int column,
+                      size_t position, const char *msg, va_list ap)
+{
+    if(!error)
+        return;
+
+    if(error->text[0] != '\0') {
+        /* error already set */
+        return;
+    }
+
+    error->line = line;
+    error->column = column;
+    error->position = position;
+
+    vsnprintf(error->text, JSON_ERROR_TEXT_LENGTH, msg, ap);
+}

src/hashtable.c

@@ -1,29 +1,28 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * This library is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
  */
 
 #include <stdlib.h>
+#include <jansson_config.h>   /* for JSON_INLINE */
+#include "jansson_private.h"  /* for container_of() */
 #include "hashtable.h"
 
 typedef struct hashtable_list list_t;
 typedef struct hashtable_pair pair_t;
 typedef struct hashtable_bucket bucket_t;
 
-#define container_of(ptr_, type_, member_)                      \
-    ((type_ *)((char *)ptr_ - (size_t)&((type_ *)0)->member_))
-
 #define list_to_pair(list_)  container_of(list_, pair_t, list)
 
-static inline void list_init(list_t *list)
+static JSON_INLINE void list_init(list_t *list)
 {
     list->next = list;
     list->prev = list;
 }
 
-static inline void list_insert(list_t *list, list_t *node)
+static JSON_INLINE void list_insert(list_t *list, list_t *node)
 {
     node->next = list;
     node->prev = list->prev;
@@ -31,13 +30,13 @@ static inline void list_insert(list_t *list, list_t *node)
     list->prev = node;
 }
 
-static inline void list_remove(list_t *list)
+static JSON_INLINE void list_remove(list_t *list)
 {
     list->prev->next = list->next;
     list->next->prev = list->prev;
 }
 
-static inline int bucket_is_empty(hashtable_t *hashtable, bucket_t *bucket)
+static JSON_INLINE int bucket_is_empty(hashtable_t *hashtable, bucket_t *bucket)
 {
     return bucket->first == &hashtable->list && bucket->first == bucket->last;
 }
@@ -57,22 +56,21 @@ static void insert_to_bucket(hashtable_t *hashtable, bucket_t *bucket,
     }
 }
 
-static unsigned int primes[] = {
+static size_t primes[] = {
     5, 13, 23, 53, 97, 193, 389, 769, 1543, 3079, 6151, 12289, 24593,
     49157, 98317, 196613, 393241, 786433, 1572869, 3145739, 6291469,
     12582917, 25165843, 50331653, 100663319, 201326611, 402653189,
     805306457, 1610612741
 };
-static const unsigned int num_primes = sizeof(primes) / sizeof(unsigned int);
 
-static inline unsigned int num_buckets(hashtable_t *hashtable)
+static JSON_INLINE size_t num_buckets(hashtable_t *hashtable)
 {
     return primes[hashtable->num_buckets];
 }
 
 
 static pair_t *hashtable_find_pair(hashtable_t *hashtable, bucket_t *bucket,
-                                   const void *key, unsigned int hash)
+                                   const void *key, size_t hash)
 {
     list_t *list;
     pair_t *pair;
@@ -98,11 +96,11 @@ static pair_t *hashtable_find_pair(hashtable_t *hashtable, bucket_t *bucket,
 
 /* returns 0 on success, -1 if key was not found */
 static int hashtable_do_del(hashtable_t *hashtable,
-                            const void *key, unsigned int hash)
+                            const void *key, size_t hash)
 {
     pair_t *pair;
     bucket_t *bucket;
-    unsigned int index;
+    size_t index;
 
     index = hash % num_buckets(hashtable);
     bucket = &hashtable->buckets[index];
@@ -127,7 +125,7 @@ static int hashtable_do_del(hashtable_t *hashtable,
     if(hashtable->free_value)
         hashtable->free_value(pair->value);
 
-    free(pair);
+    jsonp_free(pair);
     hashtable->size--;
 
     return 0;
@@ -146,7 +144,7 @@ static void hashtable_do_clear(hashtable_t *hashtable)
             hashtable->free_key(pair->key);
         if(hashtable->free_value)
             hashtable->free_value(pair->value);
-        free(pair);
+        jsonp_free(pair);
     }
 }
 
@@ -154,14 +152,14 @@ static int hashtable_do_rehash(hashtable_t *hashtable)
 {
     list_t *list, *next;
     pair_t *pair;
-    unsigned int i, index, new_size;
+    size_t i, index, new_size;
 
-    free(hashtable->buckets);
+    jsonp_free(hashtable->buckets);
 
     hashtable->num_buckets++;
     new_size = num_buckets(hashtable);
 
-    hashtable->buckets = malloc(new_size * sizeof(bucket_t));
+    hashtable->buckets = jsonp_malloc(new_size * sizeof(bucket_t));
     if(!hashtable->buckets)
         return -1;
 
@@ -188,13 +186,13 @@ static int hashtable_do_rehash(hashtable_t *hashtable)
 hashtable_t *hashtable_create(key_hash_fn hash_key, key_cmp_fn cmp_keys,
                               free_fn free_key, free_fn free_value)
 {
-    hashtable_t *hashtable = malloc(sizeof(hashtable_t));
+    hashtable_t *hashtable = jsonp_malloc(sizeof(hashtable_t));
     if(!hashtable)
         return NULL;
 
     if(hashtable_init(hashtable, hash_key, cmp_keys, free_key, free_value))
     {
-        free(hashtable);
+        jsonp_free(hashtable);
         return NULL;
     }
 
@@ -204,18 +202,18 @@ hashtable_t *hashtable_create(key_hash_fn hash_key, key_cmp_fn cmp_keys,
 void hashtable_destroy(hashtable_t *hashtable)
 {
     hashtable_close(hashtable);
-    free(hashtable);
+    jsonp_free(hashtable);
 }
 
 int hashtable_init(hashtable_t *hashtable,
                    key_hash_fn hash_key, key_cmp_fn cmp_keys,
                    free_fn free_key, free_fn free_value)
 {
-    unsigned int i;
+    size_t i;
 
     hashtable->size = 0;
     hashtable->num_buckets = 0;  /* index to primes[] */
-    hashtable->buckets = malloc(num_buckets(hashtable) * sizeof(bucket_t));
+    hashtable->buckets = jsonp_malloc(num_buckets(hashtable) * sizeof(bucket_t));
     if(!hashtable->buckets)
         return -1;
 
@@ -238,47 +236,55 @@ int hashtable_init(hashtable_t *hashtable,
 void hashtable_close(hashtable_t *hashtable)
 {
     hashtable_do_clear(hashtable);
-    free(hashtable->buckets);
+    jsonp_free(hashtable->buckets);
 }
 
 int hashtable_set(hashtable_t *hashtable, void *key, void *value)
 {
     pair_t *pair;
     bucket_t *bucket;
-    unsigned int hash, index;
-
-    hash = hashtable->hash_key(key);
-
-    /* if the key already exists, delete it */
-    hashtable_do_del(hashtable, key, hash);
+    size_t hash, index;
 
     /* rehash if the load ratio exceeds 1 */
     if(hashtable->size >= num_buckets(hashtable))
         if(hashtable_do_rehash(hashtable))
             return -1;
 
-    pair = malloc(sizeof(pair_t));
-    if(!pair)
-        return -1;
-
-    pair->key = key;
-    pair->value = value;
-    pair->hash = hash;
-    list_init(&pair->list);
-
+    hash = hashtable->hash_key(key);
     index = hash % num_buckets(hashtable);
     bucket = &hashtable->buckets[index];
+    pair = hashtable_find_pair(hashtable, bucket, key, hash);
 
-    insert_to_bucket(hashtable, bucket, &pair->list);
+    if(pair)
+    {
+        if(hashtable->free_key)
+            hashtable->free_key(key);
+        if(hashtable->free_value)
+            hashtable->free_value(pair->value);
+        pair->value = value;
+    }
+    else
+    {
+        pair = jsonp_malloc(sizeof(pair_t));
+        if(!pair)
+            return -1;
 
-    hashtable->size++;
+        pair->key = key;
+        pair->value = value;
+        pair->hash = hash;
+        list_init(&pair->list);
+
+        insert_to_bucket(hashtable, bucket, &pair->list);
+
+        hashtable->size++;
+    }
     return 0;
 }
 
 void *hashtable_get(hashtable_t *hashtable, const void *key)
 {
     pair_t *pair;
-    unsigned int hash;
+    size_t hash;
     bucket_t *bucket;
 
     hash = hashtable->hash_key(key);
@@ -293,13 +299,13 @@ void *hashtable_get(hashtable_t *hashtable, const void *key)
 
 int hashtable_del(hashtable_t *hashtable, const void *key)
 {
-    unsigned int hash = hashtable->hash_key(key);
+    size_t hash = hashtable->hash_key(key);
     return hashtable_do_del(hashtable, key, hash);
 }
 
 void hashtable_clear(hashtable_t *hashtable)
 {
-    unsigned int i;
+    size_t i;
 
     hashtable_do_clear(hashtable);
 
@@ -318,6 +324,22 @@ void *hashtable_iter(hashtable_t *hashtable)
     return hashtable_iter_next(hashtable, &hashtable->list);
 }
 
+void *hashtable_iter_at(hashtable_t *hashtable, const void *key)
+{
+    pair_t *pair;
+    size_t hash;
+    bucket_t *bucket;
+
+    hash = hashtable->hash_key(key);
+    bucket = &hashtable->buckets[hash % num_buckets(hashtable)];
+
+    pair = hashtable_find_pair(hashtable, bucket, key, hash);
+    if(!pair)
+        return NULL;
+
+    return &pair->list;
+}
+
 void *hashtable_iter_next(hashtable_t *hashtable, void *iter)
 {
     list_t *list = (list_t *)iter;
@@ -337,3 +359,13 @@ void *hashtable_iter_value(void *iter)
     pair_t *pair = list_to_pair((list_t *)iter);
     return pair->value;
 }
+
+void hashtable_iter_set(hashtable_t *hashtable, void *iter, void *value)
+{
+    pair_t *pair = list_to_pair((list_t *)iter);
+
+    if(hashtable->free_value)
+        hashtable->free_value(pair->value);
+
+    pair->value = value;
+}

src/hashtable.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * This library is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -8,7 +8,7 @@
 #ifndef HASHTABLE_H
 #define HASHTABLE_H
 
-typedef unsigned int (*key_hash_fn)(const void *key);
+typedef size_t (*key_hash_fn)(const void *key);
 typedef int (*key_cmp_fn)(const void *key1, const void *key2);
 typedef void (*free_fn)(void *key);
 
@@ -20,7 +20,7 @@ struct hashtable_list {
 struct hashtable_pair {
     void *key;
     void *value;
-    unsigned int hash;
+    size_t hash;
     struct hashtable_list list;
 };
 
@@ -30,9 +30,9 @@ struct hashtable_bucket {
 };
 
 typedef struct hashtable {
-    unsigned int size;
+    size_t size;
     struct hashtable_bucket *buckets;
-    unsigned int num_buckets;  /* index to primes[] */
+    size_t num_buckets;  /* index to primes[] */
     struct hashtable_list list;
 
     key_hash_fn hash_key;
@@ -160,6 +160,17 @@ void hashtable_clear(hashtable_t *hashtable);
  */
 void *hashtable_iter(hashtable_t *hashtable);
 
+/**
+ * hashtable_iter_at - Return an iterator at a specific key
+ *
+ * @hashtable: The hashtable object
+ * @key: The key that the iterator should point to
+ *
+ * Like hashtable_iter() but returns an iterator pointing to a
+ * specific key.
+ */
+void *hashtable_iter_at(hashtable_t *hashtable, const void *key);
+
 /**
  * hashtable_iter_next - Advance an iterator
  *
@@ -185,4 +196,12 @@ void *hashtable_iter_key(void *iter);
  */
 void *hashtable_iter_value(void *iter);
 
+/**
+ * hashtable_iter_set - Set the value pointed by an iterator
+ *
+ * @iter: The iterator
+ * @value: The value to set
+ */
+void hashtable_iter_set(hashtable_t *hashtable, void *iter, void *value);
+
 #endif

src/jansson.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -9,11 +9,31 @@
 #define JANSSON_H
 
 #include <stdio.h>
+#include <stdlib.h>  /* for size_t */
+#include <stdarg.h>
+
+#include <jansson_config.h>
 
 #ifdef __cplusplus
 extern "C" {
 #endif
 
+/* version */
+
+#define JANSSON_MAJOR_VERSION  2
+#define JANSSON_MINOR_VERSION  2
+#define JANSSON_MICRO_VERSION  1
+
+/* Micro version is omitted if it's 0 */
+#define JANSSON_VERSION  "2.2.1"
+
+/* Version as a 3-byte hex number, e.g. 0x010201 == 1.2.1. Use this
+   for numeric comparisons, e.g. #if JANSSON_VERSION_HEX >= ... */
+#define JANSSON_VERSION_HEX  ((JANSSON_MAJOR_VERSION << 16) |   \
+                              (JANSSON_MINOR_VERSION << 8)  |   \
+                              (JANSSON_MICRO_VERSION << 0))
+
+
 /* types */
 
 typedef enum {
@@ -29,9 +49,17 @@ typedef enum {
 
 typedef struct {
     json_type type;
-    unsigned long refcount;
+    size_t refcount;
 } json_t;
 
+#if JSON_INTEGER_IS_LONG_LONG
+#define JSON_INTEGER_FORMAT "lld"
+typedef long long json_int_t;
+#else
+#define JSON_INTEGER_FORMAT "ld"
+typedef long json_int_t;
+#endif /* JSON_INTEGER_IS_LONG_LONG */
+
 #define json_typeof(json)      ((json)->type)
 #define json_is_object(json)   (json && json_typeof(json) == JSON_OBJECT)
 #define json_is_array(json)    (json && json_typeof(json) == JSON_ARRAY)
@@ -50,15 +78,16 @@ json_t *json_object(void);
 json_t *json_array(void);
 json_t *json_string(const char *value);
 json_t *json_string_nocheck(const char *value);
-json_t *json_integer(int value);
+json_t *json_integer(json_int_t value);
 json_t *json_real(double value);
 json_t *json_true(void);
 json_t *json_false(void);
 json_t *json_null(void);
 
-static inline json_t *json_incref(json_t *json)
+static JSON_INLINE
+json_t *json_incref(json_t *json)
 {
-    if(json && json->refcount != (unsigned int)-1)
+    if(json && json->refcount != (size_t)-1)
         ++json->refcount;
     return json;
 }
@@ -66,16 +95,31 @@ static inline json_t *json_incref(json_t *json)
 /* do not call json_delete directly */
 void json_delete(json_t *json);
 
-static inline void json_decref(json_t *json)
+static JSON_INLINE
+void json_decref(json_t *json)
 {
-    if(json && json->refcount != (unsigned int)-1 && --json->refcount == 0)
+    if(json && json->refcount != (size_t)-1 && --json->refcount == 0)
         json_delete(json);
 }
 
 
+/* error reporting */
+
+#define JSON_ERROR_TEXT_LENGTH    160
+#define JSON_ERROR_SOURCE_LENGTH   80
+
+typedef struct {
+    int line;
+    int column;
+    int position;
+    char source[JSON_ERROR_SOURCE_LENGTH];
+    char text[JSON_ERROR_TEXT_LENGTH];
+} json_error_t;
+
+
 /* getters, setters, manipulation */
 
-unsigned int json_object_size(const json_t *object);
+size_t json_object_size(const json_t *object);
 json_t *json_object_get(const json_t *object, const char *key);
 int json_object_set_new(json_t *object, const char *key, json_t *value);
 int json_object_set_new_nocheck(json_t *object, const char *key, json_t *value);
@@ -83,60 +127,82 @@ int json_object_del(json_t *object, const char *key);
 int json_object_clear(json_t *object);
 int json_object_update(json_t *object, json_t *other);
 void *json_object_iter(json_t *object);
+void *json_object_iter_at(json_t *object, const char *key);
 void *json_object_iter_next(json_t *object, void *iter);
 const char *json_object_iter_key(void *iter);
 json_t *json_object_iter_value(void *iter);
+int json_object_iter_set_new(json_t *object, void *iter, json_t *value);
 
-static inline
+static JSON_INLINE
 int json_object_set(json_t *object, const char *key, json_t *value)
 {
     return json_object_set_new(object, key, json_incref(value));
 }
 
-static inline
+static JSON_INLINE
 int json_object_set_nocheck(json_t *object, const char *key, json_t *value)
 {
     return json_object_set_new_nocheck(object, key, json_incref(value));
 }
 
-unsigned int json_array_size(const json_t *array);
-json_t *json_array_get(const json_t *array, unsigned int index);
-int json_array_set_new(json_t *array, unsigned int index, json_t *value);
+static JSON_INLINE
+int json_object_iter_set(json_t *object, void *iter, json_t *value)
+{
+    return json_object_iter_set_new(object, iter, json_incref(value));
+}
+
+size_t json_array_size(const json_t *array);
+json_t *json_array_get(const json_t *array, size_t index);
+int json_array_set_new(json_t *array, size_t index, json_t *value);
 int json_array_append_new(json_t *array, json_t *value);
-int json_array_insert_new(json_t *array, unsigned int index, json_t *value);
-int json_array_remove(json_t *array, unsigned int index);
+int json_array_insert_new(json_t *array, size_t index, json_t *value);
+int json_array_remove(json_t *array, size_t index);
 int json_array_clear(json_t *array);
 int json_array_extend(json_t *array, json_t *other);
 
-static inline
-int json_array_set(json_t *array, unsigned int index, json_t *value)
+static JSON_INLINE
+int json_array_set(json_t *array, size_t index, json_t *value)
 {
     return json_array_set_new(array, index, json_incref(value));
 }
 
-static inline
+static JSON_INLINE
 int json_array_append(json_t *array, json_t *value)
 {
     return json_array_append_new(array, json_incref(value));
 }
 
-static inline
-int json_array_insert(json_t *array, unsigned int index, json_t *value)
+static JSON_INLINE
+int json_array_insert(json_t *array, size_t index, json_t *value)
 {
     return json_array_insert_new(array, index, json_incref(value));
 }
 
 const char *json_string_value(const json_t *string);
-int json_integer_value(const json_t *integer);
+json_int_t json_integer_value(const json_t *integer);
 double json_real_value(const json_t *real);
 double json_number_value(const json_t *json);
 
 int json_string_set(json_t *string, const char *value);
 int json_string_set_nocheck(json_t *string, const char *value);
-int json_integer_set(json_t *integer, int value);
+int json_integer_set(json_t *integer, json_int_t value);
 int json_real_set(json_t *real, double value);
 
 
+/* pack, unpack */
+
+json_t *json_pack(const char *fmt, ...);
+json_t *json_pack_ex(json_error_t *error, size_t flags, const char *fmt, ...);
+json_t *json_vpack_ex(json_error_t *error, size_t flags, const char *fmt, va_list ap);
+
+#define JSON_VALIDATE_ONLY  0x1
+#define JSON_STRICT         0x2
+
+int json_unpack(json_t *root, const char *fmt, ...);
+int json_unpack_ex(json_t *root, json_error_t *error, size_t flags, const char *fmt, ...);
+int json_vunpack_ex(json_t *root, json_error_t *error, size_t flags, const char *fmt, va_list ap);
+
+
 /* equality */
 
 int json_equal(json_t *value1, json_t *value2);
@@ -148,27 +214,39 @@ json_t *json_copy(json_t *value);
 json_t *json_deep_copy(json_t *value);
 
 
-/* loading, printing */
+/* decoding */
 
-#define JSON_ERROR_TEXT_LENGTH  160
+#define JSON_REJECT_DUPLICATES 0x1
+#define JSON_DISABLE_EOF_CHECK 0x2
 
-typedef struct {
-    char text[JSON_ERROR_TEXT_LENGTH];
-    int line;
-} json_error_t;
+json_t *json_loads(const char *input, size_t flags, json_error_t *error);
+json_t *json_loadb(const char *buffer, size_t buflen, size_t flags, json_error_t *error);
+json_t *json_loadf(FILE *input, size_t flags, json_error_t *error);
+json_t *json_load_file(const char *path, size_t flags, json_error_t *error);
 
-json_t *json_loads(const char *input, json_error_t *error);
-json_t *json_loadf(FILE *input, json_error_t *error);
-json_t *json_load_file(const char *path, json_error_t *error);
 
-#define JSON_INDENT(n)      (n & 0xFF)
-#define JSON_COMPACT        0x100
-#define JSON_ENSURE_ASCII   0x200
-#define JSON_SORT_KEYS      0x400
+/* encoding */
 
-char *json_dumps(const json_t *json, unsigned long flags);
-int json_dumpf(const json_t *json, FILE *output, unsigned long flags);
-int json_dump_file(const json_t *json, const char *path, unsigned long flags);
+#define JSON_INDENT(n)      (n & 0x1F)
+#define JSON_COMPACT        0x20
+#define JSON_ENSURE_ASCII   0x40
+#define JSON_SORT_KEYS      0x80
+#define JSON_PRESERVE_ORDER 0x100
+#define JSON_ENCODE_ANY     0x200
+
+typedef int (*json_dump_callback_t)(const char *buffer, size_t size, void *data);
+
+char *json_dumps(const json_t *json, size_t flags);
+int json_dumpf(const json_t *json, FILE *output, size_t flags);
+int json_dump_file(const json_t *json, const char *path, size_t flags);
+int json_dump_callback(const json_t *json, json_dump_callback_t callback, void *data, size_t flags);
+
+/* custom memory allocation */
+
+typedef void *(*json_malloc_t)(size_t);
+typedef void (*json_free_t)(void *);
+
+void json_set_alloc_funcs(json_malloc_t malloc_fn, json_free_t free_fn);
 
 #ifdef __cplusplus
 }

src/jansson_config.h.in

@@ -0,0 +1,38 @@
+/*
+ * Copyright (c) 2010-2011 Petri Lehtinen <petri@digip.org>
+ *
+ * Jansson is free software; you can redistribute it and/or modify
+ * it under the terms of the MIT license. See LICENSE for details.
+ *
+ *
+ * This file specifies a part of the site-specific configuration for
+ * Jansson, namely those things that affect the public API in
+ * jansson.h.
+ *
+ * The configure script copies this file to jansson_config.h and
+ * replaces @var@ substitutions by values that fit your system. If you
+ * cannot run the configure script, you can do the value substitution
+ * by hand.
+ */
+
+#ifndef JANSSON_CONFIG_H
+#define JANSSON_CONFIG_H
+
+/* If your compiler supports the inline keyword in C, JSON_INLINE is
+   defined to `inline', otherwise empty. In C++, the inline is always
+   supported. */
+#ifdef __cplusplus
+#define JSON_INLINE inline
+#else
+#define JSON_INLINE @json_inline@
+#endif
+
+/* If your compiler supports the `long long` type,
+   JSON_INTEGER_IS_LONG_LONG is defined to 1, otherwise to 0. */
+#define JSON_INTEGER_IS_LONG_LONG @json_have_long_long@
+
+/* If locale.h and localeconv() are available, define to 1,
+   otherwise to 0. */
+#define JSON_HAVE_LOCALECONV @json_have_localeconv@
+
+#endif

src/jansson_config.h.win32

@@ -0,0 +1,38 @@
+/*
+ * Copyright (c) 2010-2011 Petri Lehtinen <petri@digip.org>
+ *
+ * Jansson is free software; you can redistribute it and/or modify
+ * it under the terms of the MIT license. See LICENSE for details.
+ *
+ *
+ * This file specifies a part of the site-specific configuration for
+ * Jansson, namely those things that affect the public API in
+ * jansson.h.
+ *
+ * The configure script copies this file to jansson_config.h and
+ * replaces @var@ substitutions by values that fit your system. If you
+ * cannot run the configure script, you can do the value substitution
+ * by hand.
+ */
+
+#ifndef JANSSON_CONFIG_H
+#define JANSSON_CONFIG_H
+
+/* If your compiler supports the inline keyword in C, JSON_INLINE is
+   defined to `inline', otherwise empty. In C++, the inline is always
+   supported. */
+#ifdef __cplusplus
+#define JSON_INLINE inline
+#else
+#define JSON_INLINE 
+#endif
+
+/* If your compiler supports the `long long` type,
+   JSON_INTEGER_IS_LONG_LONG is defined to 1, otherwise to 0. */
+#define JSON_INTEGER_IS_LONG_LONG 1
+
+/* If locale.h and localeconv() are available, define to 1,
+   otherwise to 0. */
+#define JSON_HAVE_LOCALECONV 1
+
+#endif

src/jansson_private.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -8,22 +8,40 @@
 #ifndef JANSSON_PRIVATE_H
 #define JANSSON_PRIVATE_H
 
+#include <stddef.h>
 #include "jansson.h"
 #include "hashtable.h"
+#include "strbuffer.h"
 
 #define container_of(ptr_, type_, member_)  \
-    ((type_ *)((char *)ptr_ - (size_t)&((type_ *)0)->member_))
+    ((type_ *)((char *)ptr_ - offsetof(type_, member_)))
+
+/* On some platforms, max() may already be defined */
+#ifndef max
+#define max(a, b)  ((a) > (b) ? (a) : (b))
+#endif
+
+/* va_copy is a C99 feature. In C89 implementations, it's sometimes
+   available as __va_copy. If not, memcpy() should do the trick. */
+#ifndef va_copy
+#ifdef __va_copy
+#define va_copy __va_copy
+#else
+#define va_copy(a, b)  memcpy(&(a), &(b), sizeof(va_list))
+#endif
+#endif
 
 typedef struct {
     json_t json;
     hashtable_t hashtable;
+    size_t serial;
     int visited;
 } json_object_t;
 
 typedef struct {
     json_t json;
-    unsigned int size;
-    unsigned int entries;
+    size_t size;
+    size_t entries;
     json_t **table;
     int visited;
 } json_array_t;
@@ -40,7 +58,7 @@ typedef struct {
 
 typedef struct {
     json_t json;
-    int value;
+    json_int_t value;
 } json_integer_t;
 
 #define json_to_object(json_)  container_of(json_, json_object_t, json)
@@ -49,4 +67,30 @@ typedef struct {
 #define json_to_real(json_)   container_of(json_, json_real_t, json)
 #define json_to_integer(json_) container_of(json_, json_integer_t, json)
 
+size_t jsonp_hash_str(const void *ptr);
+int jsonp_str_equal(const void *ptr1, const void *ptr2);
+
+typedef struct {
+    size_t serial;
+    char key[1];
+} object_key_t;
+
+const object_key_t *jsonp_object_iter_fullkey(void *iter);
+
+void jsonp_error_init(json_error_t *error, const char *source);
+void jsonp_error_set_source(json_error_t *error, const char *source);
+void jsonp_error_set(json_error_t *error, int line, int column,
+                     size_t position, const char *msg, ...);
+void jsonp_error_vset(json_error_t *error, int line, int column,
+                      size_t position, const char *msg, va_list ap);
+
+/* Locale independent string<->double conversions */
+int jsonp_strtod(strbuffer_t *strbuffer, double *out);
+int jsonp_dtostr(char *buffer, size_t size, double value);
+
+/* Wrappers for custom memory functions */
+void* jsonp_malloc(size_t size);
+void jsonp_free(void *ptr);
+char *jsonp_strdup(const char *str);
+
 #endif

src/load.c

@@ -1,26 +1,27 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
  */
 
 #define _GNU_SOURCE
-#include <ctype.h>
 #include <errno.h>
 #include <limits.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#include <stdarg.h>
 #include <assert.h>
-#include <stdint.h>
 
 #include <jansson.h>
 #include "jansson_private.h"
 #include "strbuffer.h"
 #include "utf.h"
 
+#define STREAM_STATE_OK        0
+#define STREAM_STATE_EOF      -1
+#define STREAM_STATE_ERROR    -2
+
 #define TOKEN_INVALID         -1
 #define TOKEN_EOF              0
 #define TOKEN_STRING         256
@@ -30,112 +31,133 @@
 #define TOKEN_FALSE          260
 #define TOKEN_NULL           261
 
-/* read one byte from stream, return EOF on end of file */
-typedef int (*get_func)(void *data);
+/* Locale independent versions of isxxx() functions */
+#define l_isupper(c)  ('A' <= c && c <= 'Z')
+#define l_islower(c)  ('a' <= c && c <= 'z')
+#define l_isalpha(c)  (l_isupper(c) || l_islower(c))
+#define l_isdigit(c)  ('0' <= c && c <= '9')
+#define l_isxdigit(c) \
+    (l_isdigit(c) || 'A' <= c || c <= 'F' || 'a' <= c || c <= 'f')
 
-/* return non-zero if end of file has been reached */
-typedef int (*eof_func)(void *data);
+/* Read one byte from stream, convert to unsigned char, then int, and
+   return. return EOF on end of file. This corresponds to the
+   behaviour of fgetc(). */
+typedef int (*get_func)(void *data);
 
 typedef struct {
     get_func get;
-    eof_func eof;
     void *data;
-    int stream_pos;
     char buffer[5];
     int buffer_pos;
+    int state;
+    int line;
+    int column, last_column;
+    size_t position;
 } stream_t;
 
-
 typedef struct {
     stream_t stream;
     strbuffer_t saved_text;
     int token;
-    int line, column;
     union {
         char *string;
-        int integer;
+        json_int_t integer;
         double real;
     } value;
 } lex_t;
 
+#define stream_to_lex(stream) container_of(stream, lex_t, stream)
+
 
 /*** error reporting ***/
 
-static void error_init(json_error_t *error)
-{
-    if(error)
-    {
-        error->text[0] = '\0';
-        error->line = -1;
-    }
-}
-
 static void error_set(json_error_t *error, const lex_t *lex,
                       const char *msg, ...)
 {
     va_list ap;
-    char text[JSON_ERROR_TEXT_LENGTH];
+    char msg_text[JSON_ERROR_TEXT_LENGTH];
+    char msg_with_context[JSON_ERROR_TEXT_LENGTH];
 
-    if(!error || error->text[0] != '\0') {
-        /* error already set */
+    int line = -1, col = -1;
+    size_t pos = 0;
+    const char *result = msg_text;
+
+    if(!error)
         return;
-    }
 
     va_start(ap, msg);
-    vsnprintf(text, JSON_ERROR_TEXT_LENGTH, msg, ap);
+    vsnprintf(msg_text, JSON_ERROR_TEXT_LENGTH, msg, ap);
     va_end(ap);
 
     if(lex)
     {
         const char *saved_text = strbuffer_value(&lex->saved_text);
-        error->line = lex->line;
+
+        line = lex->stream.line;
+        col = lex->stream.column;
+        pos = lex->stream.position;
+
         if(saved_text && saved_text[0])
         {
             if(lex->saved_text.length <= 20) {
-                snprintf(error->text, JSON_ERROR_TEXT_LENGTH,
-                         "%s near '%s'", text, saved_text);
+                snprintf(msg_with_context, JSON_ERROR_TEXT_LENGTH,
+                         "%s near '%s'", msg_text, saved_text);
+                result = msg_with_context;
             }
-            else
-                snprintf(error->text, JSON_ERROR_TEXT_LENGTH, "%s", text);
         }
         else
         {
-            snprintf(error->text, JSON_ERROR_TEXT_LENGTH,
-                     "%s near end of file", text);
+            if(lex->stream.state == STREAM_STATE_ERROR) {
+                /* No context for UTF-8 decoding errors */
+                result = msg_text;
+            }
+            else {
+                snprintf(msg_with_context, JSON_ERROR_TEXT_LENGTH,
+                         "%s near end of file", msg_text);
+                result = msg_with_context;
+            }
         }
     }
-    else
-    {
-        error->line = -1;
-        snprintf(error->text, JSON_ERROR_TEXT_LENGTH, "%s", text);
-    }
+
+    jsonp_error_set(error, line, col, pos, "%s", result);
 }
 
 
 /*** lexical analyzer ***/
 
-void stream_init(stream_t *stream, get_func get, eof_func eof, void *data)
+static void
+stream_init(stream_t *stream, get_func get, void *data)
 {
     stream->get = get;
-    stream->eof = eof;
     stream->data = data;
-    stream->stream_pos = 0;
     stream->buffer[0] = '\0';
     stream->buffer_pos = 0;
+
+    stream->state = STREAM_STATE_OK;
+    stream->line = 1;
+    stream->column = 0;
+    stream->position = 0;
 }
 
-static char stream_get(stream_t *stream, json_error_t *error)
+static int stream_get(stream_t *stream, json_error_t *error)
 {
-    char c;
+    int c;
+
+    if(stream->state != STREAM_STATE_OK)
+        return stream->state;
 
     if(!stream->buffer[stream->buffer_pos])
     {
-        stream->buffer[0] = stream->get(stream->data);
+        c = stream->get(stream->data);
+        if(c == EOF) {
+            stream->state = STREAM_STATE_EOF;
+            return STREAM_STATE_EOF;
+        }
+
+        stream->buffer[0] = c;
         stream->buffer_pos = 0;
 
-        c = stream->buffer[0];
-
-        if((unsigned char)c >= 0x80 && c != (char)EOF)
+        if(0x80 <= c && c <= 0xFF)
         {
             /* multi-byte UTF-8 sequence */
             int i, count;
@@ -152,30 +174,47 @@ static char stream_get(stream_t *stream, json_error_t *error)
             if(!utf8_check_full(stream->buffer, count, NULL))
                 goto out;
 
-            stream->stream_pos += count;
             stream->buffer[count] = '\0';
         }
-        else {
+        else
             stream->buffer[1] = '\0';
-            stream->stream_pos++;
-        }
     }
 
-    return stream->buffer[stream->buffer_pos++];
+    c = stream->buffer[stream->buffer_pos++];
+
+    stream->position++;
+    if(c == '\n') {
+        stream->line++;
+        stream->last_column = stream->column;
+        stream->column = 0;
+    }
+    else if(utf8_check_first(c)) {
+        /* track the Unicode character column, so increment only if
+           this is the first character of a UTF-8 sequence */
+        stream->column++;
+    }
+
+    return c;
 
 out:
-    error_set(error, NULL, "unable to decode byte 0x%x at position %d",
-              (unsigned char)c, stream->stream_pos);
-
-    stream->buffer[0] = EOF;
-    stream->buffer[1] = '\0';
-    stream->buffer_pos = 1;
-
-    return EOF;
+    stream->state = STREAM_STATE_ERROR;
+    error_set(error, stream_to_lex(stream), "unable to decode byte 0x%x", c);
+    return STREAM_STATE_ERROR;
 }
 
-static void stream_unget(stream_t *stream, char c)
+static void stream_unget(stream_t *stream, int c)
 {
+    if(c == STREAM_STATE_EOF || c == STREAM_STATE_ERROR)
+        return;
+
+    stream->position--;
+    if(c == '\n') {
+        stream->line--;
+        stream->column = stream->last_column;
+    }
+    else if(utf8_check_first(c))
+        stream->column--;
+
     assert(stream->buffer_pos > 0);
     stream->buffer_pos--;
     assert(stream->buffer[stream->buffer_pos] == c);
@@ -187,29 +226,32 @@ static int lex_get(lex_t *lex, json_error_t *error)
     return stream_get(&lex->stream, error);
 }
 
-static int lex_eof(lex_t *lex)
-{
-    return lex->stream.eof(lex->stream.data);
-}
-
-static void lex_save(lex_t *lex, char c)
+static void lex_save(lex_t *lex, int c)
 {
     strbuffer_append_byte(&lex->saved_text, c);
 }
 
 static int lex_get_save(lex_t *lex, json_error_t *error)
 {
-    char c = stream_get(&lex->stream, error);
-    lex_save(lex, c);
+    int c = stream_get(&lex->stream, error);
+    if(c != STREAM_STATE_EOF && c != STREAM_STATE_ERROR)
+        lex_save(lex, c);
     return c;
 }
 
-static void lex_unget_unsave(lex_t *lex, char c)
+static void lex_unget(lex_t *lex, int c)
 {
-    char d;
     stream_unget(&lex->stream, c);
-    d = strbuffer_pop(&lex->saved_text);
-    assert(c == d);
+}
+
+static void lex_unget_unsave(lex_t *lex, int c)
+{
+    if(c != STREAM_STATE_EOF && c != STREAM_STATE_ERROR) {
+        char d;
+        stream_unget(&lex->stream, c);
+        d = strbuffer_pop(&lex->saved_text);
+        assert(c == d);
+    }
 }
 
 static void lex_save_cached(lex_t *lex)
@@ -218,6 +260,7 @@ static void lex_save_cached(lex_t *lex)
     {
         lex_save(lex, lex->stream.buffer[lex->stream.buffer_pos]);
         lex->stream.buffer_pos++;
+        lex->stream.position++;
     }
 }
 
@@ -232,11 +275,11 @@ static int32_t decode_unicode_escape(const char *str)
     for(i = 1; i <= 4; i++) {
         char c = str[i];
         value <<= 4;
-        if(isdigit(c))
+        if(l_isdigit(c))
             value += c - '0';
-        else if(islower(c))
+        else if(l_islower(c))
             value += c - 'a' + 10;
-        else if(isupper(c))
+        else if(l_isupper(c))
             value += c - 'A' + 10;
         else
             assert(0);
@@ -247,7 +290,7 @@ static int32_t decode_unicode_escape(const char *str)
 
 static void lex_scan_string(lex_t *lex, json_error_t *error)
 {
-    char c;
+    int c;
     const char *p;
     char *t;
     int i;
@@ -258,14 +301,15 @@ static void lex_scan_string(lex_t *lex, json_error_t *error)
     c = lex_get_save(lex, error);
 
     while(c != '"') {
-        if(c == (char)EOF) {
-            lex_unget_unsave(lex, c);
-            if(lex_eof(lex))
-                error_set(error, lex, "premature end of input");
+        if(c == STREAM_STATE_ERROR)
+            goto out;
+
+        else if(c == STREAM_STATE_EOF) {
+            error_set(error, lex, "premature end of input");
             goto out;
         }
 
-        else if((unsigned char)c <= 0x1F) {
+        else if(0 <= c && c <= 0x1F) {
             /* control character */
             lex_unget_unsave(lex, c);
             if(c == '\n')
@@ -280,8 +324,7 @@ static void lex_scan_string(lex_t *lex, json_error_t *error)
             if(c == 'u') {
                 c = lex_get_save(lex, error);
                 for(i = 0; i < 4; i++) {
-                    if(!isxdigit(c)) {
-                        lex_unget_unsave(lex, c);
+                    if(!l_isxdigit(c)) {
                         error_set(error, lex, "invalid escape");
                         goto out;
                     }
@@ -292,7 +335,6 @@ static void lex_scan_string(lex_t *lex, json_error_t *error)
                     c == 'f' || c == 'n' || c == 'r' || c == 't')
                 c = lex_get_save(lex, error);
             else {
-                lex_unget_unsave(lex, c);
                 error_set(error, lex, "invalid escape");
                 goto out;
             }
@@ -308,7 +350,7 @@ static void lex_scan_string(lex_t *lex, json_error_t *error)
          - two \uXXXX escapes (length 12) forming an UTF-16 surrogate pair
            are converted to 4 bytes
     */
-    lex->value.string = malloc(lex->saved_text.length + 1);
+    lex->value.string = jsonp_malloc(lex->saved_text.length + 1);
     if(!lex->value.string) {
         /* this is not very nice, since TOKEN_INVALID is returned */
         goto out;
@@ -398,10 +440,16 @@ static void lex_scan_string(lex_t *lex, json_error_t *error)
     return;
 
 out:
-    free(lex->value.string);
+    jsonp_free(lex->value.string);
 }
 
-static int lex_scan_number(lex_t *lex, char c, json_error_t *error)
+#if JSON_INTEGER_IS_LONG_LONG
+#define json_strtoint     strtoll
+#else
+#define json_strtoint     strtol
+#endif
+
+static int lex_scan_number(lex_t *lex, int c, json_error_t *error)
 {
     const char *saved_text;
     char *end;
@@ -414,52 +462,55 @@ static int lex_scan_number(lex_t *lex, char c, json_error_t *error)
 
     if(c == '0') {
         c = lex_get_save(lex, error);
-        if(isdigit(c)) {
+        if(l_isdigit(c)) {
             lex_unget_unsave(lex, c);
             goto out;
         }
     }
-    else if(isdigit(c)) {
+    else if(l_isdigit(c)) {
         c = lex_get_save(lex, error);
-        while(isdigit(c))
+        while(l_isdigit(c))
             c = lex_get_save(lex, error);
     }
     else {
-      lex_unget_unsave(lex, c);
-      goto out;
+        lex_unget_unsave(lex, c);
+        goto out;
     }
 
     if(c != '.' && c != 'E' && c != 'e') {
-        long value;
+        json_int_t value;
 
         lex_unget_unsave(lex, c);
 
         saved_text = strbuffer_value(&lex->saved_text);
-        value = strtol(saved_text, &end, 10);
+
+        errno = 0;
+        value = json_strtoint(saved_text, &end, 10);
+        if(errno == ERANGE) {
+            if(value < 0)
+                error_set(error, lex, "too big negative integer");
+            else
+                error_set(error, lex, "too big integer");
+            goto out;
+        }
+
         assert(end == saved_text + lex->saved_text.length);
 
-        if((value == LONG_MAX && errno == ERANGE) || value > INT_MAX) {
-            error_set(error, lex, "too big integer");
-            goto out;
-        }
-        else if((value == LONG_MIN && errno == ERANGE) || value < INT_MIN) {
-            error_set(error, lex, "too big negative integer");
-            goto out;
-        }
-
         lex->token = TOKEN_INTEGER;
-        lex->value.integer = (int)value;
+        lex->value.integer = value;
         return 0;
     }
 
     if(c == '.') {
         c = lex_get(lex, error);
-        if(!isdigit(c))
+        if(!l_isdigit(c)) {
+            lex_unget(lex, c);
             goto out;
+        }
         lex_save(lex, c);
 
         c = lex_get_save(lex, error);
-        while(isdigit(c))
+        while(l_isdigit(c))
             c = lex_get_save(lex, error);
     }
 
@@ -468,30 +519,19 @@ static int lex_scan_number(lex_t *lex, char c, json_error_t *error)
         if(c == '+' || c == '-')
             c = lex_get_save(lex, error);
 
-        if(!isdigit(c)) {
+        if(!l_isdigit(c)) {
             lex_unget_unsave(lex, c);
             goto out;
         }
 
         c = lex_get_save(lex, error);
-        while(isdigit(c))
+        while(l_isdigit(c))
             c = lex_get_save(lex, error);
     }
 
     lex_unget_unsave(lex, c);
 
-    saved_text = strbuffer_value(&lex->saved_text);
-    value = strtod(saved_text, &end);
-    assert(end == saved_text + lex->saved_text.length);
-
-    if(value == 0 && errno == ERANGE) {
-        error_set(error, lex, "real number underflow");
-        goto out;
-    }
-
-    /* Cannot test for +/-HUGE_VAL because the HUGE_VAL constant is
-       only defined in C99 mode. So let's trust in sole errno. */
-    else if(errno == ERANGE) {
+    if(jsonp_strtod(&lex->saved_text, &value)) {
         error_set(error, lex, "real number overflow");
         goto out;
     }
@@ -506,29 +546,26 @@ out:
 
 static int lex_scan(lex_t *lex, json_error_t *error)
 {
-    char c;
+    int c;
 
     strbuffer_clear(&lex->saved_text);
 
     if(lex->token == TOKEN_STRING) {
-        free(lex->value.string);
+        jsonp_free(lex->value.string);
         lex->value.string = NULL;
     }
 
     c = lex_get(lex, error);
     while(c == ' ' || c == '\t' || c == '\n' || c == '\r')
-    {
-        if(c == '\n')
-            lex->line++;
-
         c = lex_get(lex, error);
+
+    if(c == STREAM_STATE_EOF) {
+        lex->token = TOKEN_EOF;
+        goto out;
     }
 
-    if(c == (char)EOF) {
-        if(lex_eof(lex))
-            lex->token = TOKEN_EOF;
-        else
-            lex->token = TOKEN_INVALID;
+    if(c == STREAM_STATE_ERROR) {
+        lex->token = TOKEN_INVALID;
         goto out;
     }
 
@@ -540,17 +577,17 @@ static int lex_scan(lex_t *lex, json_error_t *error)
     else if(c == '"')
         lex_scan_string(lex, error);
 
-    else if(isdigit(c) || c == '-') {
+    else if(l_isdigit(c) || c == '-') {
         if(lex_scan_number(lex, c, error))
             goto out;
     }
 
-    else if(isupper(c) || islower(c)) {
+    else if(l_isalpha(c)) {
         /* eat up the whole identifier for clearer error messages */
         const char *saved_text;
 
         c = lex_get_save(lex, error);
-        while(isupper(c) || islower(c))
+        while(l_isalpha(c))
             c = lex_get_save(lex, error);
         lex_unget_unsave(lex, c);
 
@@ -588,31 +625,29 @@ static char *lex_steal_string(lex_t *lex)
     return result;
 }
 
-static int lex_init(lex_t *lex, get_func get, eof_func eof, void *data)
+static int lex_init(lex_t *lex, get_func get, void *data)
 {
-    stream_init(&lex->stream, get, eof, data);
+    stream_init(&lex->stream, get, data);
     if(strbuffer_init(&lex->saved_text))
         return -1;
 
     lex->token = TOKEN_INVALID;
-    lex->line = 1;
-
     return 0;
 }
 
 static void lex_close(lex_t *lex)
 {
     if(lex->token == TOKEN_STRING)
-        free(lex->value.string);
+        jsonp_free(lex->value.string);
     strbuffer_close(&lex->saved_text);
 }
 
 
 /*** parser ***/
 
-static json_t *parse_value(lex_t *lex, json_error_t *error);
+static json_t *parse_value(lex_t *lex, size_t flags, json_error_t *error);
 
-static json_t *parse_object(lex_t *lex, json_error_t *error)
+static json_t *parse_object(lex_t *lex, size_t flags, json_error_t *error)
 {
     json_t *object = json_object();
     if(!object)
@@ -635,28 +670,36 @@ static json_t *parse_object(lex_t *lex, json_error_t *error)
         if(!key)
             return NULL;
 
+        if(flags & JSON_REJECT_DUPLICATES) {
+            if(json_object_get(object, key)) {
+                jsonp_free(key);
+                error_set(error, lex, "duplicate object key");
+                goto error;
+            }
+        }
+
         lex_scan(lex, error);
         if(lex->token != ':') {
-            free(key);
+            jsonp_free(key);
             error_set(error, lex, "':' expected");
             goto error;
         }
 
         lex_scan(lex, error);
-        value = parse_value(lex, error);
+        value = parse_value(lex, flags, error);
         if(!value) {
-            free(key);
+            jsonp_free(key);
             goto error;
         }
 
         if(json_object_set_nocheck(object, key, value)) {
-            free(key);
+            jsonp_free(key);
             json_decref(value);
             goto error;
         }
 
         json_decref(value);
-        free(key);
+        jsonp_free(key);
 
         lex_scan(lex, error);
         if(lex->token != ',')
@@ -677,7 +720,7 @@ error:
     return NULL;
 }
 
-static json_t *parse_array(lex_t *lex, json_error_t *error)
+static json_t *parse_array(lex_t *lex, size_t flags, json_error_t *error)
 {
     json_t *array = json_array();
     if(!array)
@@ -688,7 +731,7 @@ static json_t *parse_array(lex_t *lex, json_error_t *error)
         return array;
 
     while(lex->token) {
-        json_t *elem = parse_value(lex, error);
+        json_t *elem = parse_value(lex, flags, error);
         if(!elem)
             goto error;
 
@@ -717,7 +760,7 @@ error:
     return NULL;
 }
 
-static json_t *parse_value(lex_t *lex, json_error_t *error)
+static json_t *parse_value(lex_t *lex, size_t flags, json_error_t *error)
 {
     json_t *json;
 
@@ -750,11 +793,11 @@ static json_t *parse_value(lex_t *lex, json_error_t *error)
             break;
 
         case '{':
-            json = parse_object(lex, error);
+            json = parse_object(lex, flags, error);
             break;
 
         case '[':
-            json = parse_array(lex, error);
+            json = parse_array(lex, flags, error);
             break;
 
         case TOKEN_INVALID:
@@ -772,9 +815,9 @@ static json_t *parse_value(lex_t *lex, json_error_t *error)
     return json;
 }
 
-static json_t *parse_json(lex_t *lex, json_error_t *error)
+static json_t *parse_json(lex_t *lex, size_t flags, json_error_t *error)
 {
-    error_init(error);
+    json_t *result;
 
     lex_scan(lex, error);
     if(lex->token != '[' && lex->token != '{') {
@@ -782,7 +825,20 @@ static json_t *parse_json(lex_t *lex, json_error_t *error)
         return NULL;
     }
 
-    return parse_value(lex, error);
+    result = parse_value(lex, flags, error);
+    if(!result)
+        return NULL;
+
+    if(!(flags & JSON_DISABLE_EOF_CHECK)) {
+        lex_scan(lex, error);
+        if(lex->token != TOKEN_EOF) {
+            error_set(error, lex, "end of file expected");
+            json_decref(result);
+            result = NULL;
+        }
+    }
+
+    return result;
 }
 
 typedef struct
@@ -801,77 +857,97 @@ static int string_get(void *data)
     else
     {
         stream->pos++;
-        return c;
+        return (unsigned char)c;
     }
 }
 
-static int string_eof(void *data)
-{
-    string_data_t *stream = (string_data_t *)data;
-    return (stream->data[stream->pos] == '\0');
-}
-
-json_t *json_loads(const char *string, json_error_t *error)
+json_t *json_loads(const char *string, size_t flags, json_error_t *error)
 {
     lex_t lex;
     json_t *result;
+    string_data_t stream_data;
 
-    string_data_t stream_data = {
-        .data = string,
-        .pos = 0
-    };
+    stream_data.data = string;
+    stream_data.pos = 0;
 
-    if(lex_init(&lex, string_get, string_eof, (void *)&stream_data))
+    if(lex_init(&lex, string_get, (void *)&stream_data))
         return NULL;
 
-    result = parse_json(&lex, error);
-    if(!result)
-        goto out;
+    jsonp_error_init(error, "<string>");
+    result = parse_json(&lex, flags, error);
 
-    lex_scan(&lex, error);
-    if(lex.token != TOKEN_EOF) {
-        error_set(error, &lex, "end of file expected");
-        json_decref(result);
-        result = NULL;
-    }
-
-out:
     lex_close(&lex);
     return result;
 }
 
-json_t *json_loadf(FILE *input, json_error_t *error)
+typedef struct
+{
+    const char *data;
+    size_t len;
+    size_t pos;
+} buffer_data_t;
+
+static int buffer_get(void *data)
+{
+    char c;
+    buffer_data_t *stream = data;
+    if(stream->pos >= stream->len)
+      return EOF;
+
+    c = stream->data[stream->pos];
+    stream->pos++;
+    return (unsigned char)c;
+}
+
+json_t *json_loadb(const char *buffer, size_t buflen, size_t flags, json_error_t *error)
 {
     lex_t lex;
     json_t *result;
+    buffer_data_t stream_data;
 
-    if(lex_init(&lex, (get_func)fgetc, (eof_func)feof, input))
+    stream_data.data = buffer;
+    stream_data.pos = 0;
+    stream_data.len = buflen;
+
+    if(lex_init(&lex, buffer_get, (void *)&stream_data))
         return NULL;
 
-    result = parse_json(&lex, error);
-    if(!result)
-        goto out;
+    jsonp_error_init(error, "<buffer>");
+    result = parse_json(&lex, flags, error);
 
-    lex_scan(&lex, error);
-    if(lex.token != TOKEN_EOF) {
-        error_set(error, &lex, "end of file expected");
-        json_decref(result);
-        result = NULL;
-    }
-
-out:
     lex_close(&lex);
     return result;
 }
 
-json_t *json_load_file(const char *path, json_error_t *error)
+json_t *json_loadf(FILE *input, size_t flags, json_error_t *error)
+{
+    lex_t lex;
+    const char *source;
+    json_t *result;
+
+    if(lex_init(&lex, (get_func)fgetc, input))
+        return NULL;
+
+    if(input == stdin)
+        source = "<stdin>";
+    else
+        source = "<stream>";
+
+    jsonp_error_init(error, source);
+    result = parse_json(&lex, flags, error);
+
+    lex_close(&lex);
+    return result;
+}
+
+json_t *json_load_file(const char *path, size_t flags, json_error_t *error)
 {
     json_t *result;
     FILE *fp;
 
-    error_init(error);
+    jsonp_error_init(error, path);
 
-    fp = fopen(path, "r");
+    fp = fopen(path, "rb");
     if(!fp)
     {
         error_set(error, NULL, "unable to open %s: %s",
@@ -879,7 +955,7 @@ json_t *json_load_file(const char *path, json_error_t *error)
         return NULL;
     }
 
-    result = json_loadf(fp, error);
+    result = json_loadf(fp, flags, error);
 
     fclose(fp);
     return result;

src/memory.c

@@ -0,0 +1,51 @@
+/*
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2011 Basile Starynkevitch  <basile@starynkevitch.net>
+ *
+ * Jansson is free software; you can redistribute it and/or modify it
+ * under the terms of the MIT license. See LICENSE for details.
+ */
+
+#include <stdlib.h>
+#include <string.h>
+
+#include <jansson.h>
+#include "jansson_private.h"
+
+/* memory function pointers */
+static json_malloc_t do_malloc = malloc;
+static json_free_t do_free = free;
+
+void *jsonp_malloc(size_t size)
+{
+    if(!size)
+        return NULL;
+
+    return (*do_malloc)(size);
+}
+
+void jsonp_free(void *ptr)
+{
+    if(!ptr)
+        return;
+
+    (*do_free)(ptr);
+}
+
+char *jsonp_strdup(const char *str)
+{
+    char *new_str;
+
+    new_str = jsonp_malloc(strlen(str) + 1);
+    if(!new_str)
+        return NULL;
+
+    strcpy(new_str, str);
+    return new_str;
+}
+
+void json_set_alloc_funcs(json_malloc_t malloc_fn, json_free_t free_fn)
+{
+    do_malloc = malloc_fn;
+    do_free = free_fn;
+}

src/pack_unpack.c

@@ -0,0 +1,610 @@
+/*
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2011 Graeme Smecher <graeme.smecher@mail.mcgill.ca>
+ *
+ * Jansson is free software; you can redistribute it and/or modify
+ * it under the terms of the MIT license. See LICENSE for details.
+ */
+
+#include <string.h>
+#include <jansson.h>
+#include "jansson_private.h"
+#include "utf.h"
+
+typedef struct {
+    const char *start;
+    const char *fmt;
+    char token;
+    json_error_t *error;
+    size_t flags;
+    int line;
+    int column;
+} scanner_t;
+
+static const char *type_names[] = {
+    "object",
+    "array",
+    "string",
+    "integer",
+    "real",
+    "true",
+    "false",
+    "null"
+};
+
+#define type_name(x) type_names[json_typeof(x)]
+
+static const char *unpack_value_starters = "{[siIbfFOon";
+
+
+static void scanner_init(scanner_t *s, json_error_t *error,
+                         size_t flags, const char *fmt)
+{
+    s->error = error;
+    s->flags = flags;
+    s->fmt = s->start = fmt;
+    s->line = 1;
+    s->column = 0;
+}
+
+static void next_token(scanner_t *s)
+{
+    const char *t = s->fmt;
+    s->column++;
+
+    /* skip space and ignored chars */
+    while(*t == ' ' || *t == '\t' || *t == '\n' || *t == ',' || *t == ':') {
+        if(*t == '\n') {
+            s->line++;
+            s->column = 1;
+        }
+        else
+            s->column++;
+
+        t++;
+    }
+
+    s->token = *t;
+
+    t++;
+    s->fmt = t;
+}
+
+static void set_error(scanner_t *s, const char *source, const char *fmt, ...)
+{
+    va_list ap;
+    size_t pos;
+    va_start(ap, fmt);
+
+    pos = (size_t)(s->fmt - s->start);
+    jsonp_error_vset(s->error, s->line, s->column, pos, fmt, ap);
+
+    jsonp_error_set_source(s->error, source);
+
+    va_end(ap);
+}
+
+static json_t *pack(scanner_t *s, va_list *ap);
+
+static json_t *pack_object(scanner_t *s, va_list *ap)
+{
+    json_t *object = json_object();
+    next_token(s);
+
+    while(s->token != '}') {
+        const char *key;
+        json_t *value;
+
+        if(!s->token) {
+            set_error(s, "<format>", "Unexpected end of format string");
+            goto error;
+        }
+
+        if(s->token != 's') {
+            set_error(s, "<format>", "Expected format 's', got '%c'", s->token);
+            goto error;
+        }
+
+        key = va_arg(*ap, const char *);
+        if(!key) {
+            set_error(s, "<args>", "NULL object key");
+            goto error;
+        }
+
+        if(!utf8_check_string(key, -1)) {
+            set_error(s, "<args>", "Invalid UTF-8 in object key");
+            goto error;
+        }
+
+        next_token(s);
+
+        value = pack(s, ap);
+        if(!value)
+            goto error;
+
+        if(json_object_set_new_nocheck(object, key, value)) {
+            set_error(s, "<internal>", "Unable to add key \"%s\"", key);
+            goto error;
+        }
+
+        next_token(s);
+    }
+
+    return object;
+
+error:
+    json_decref(object);
+    return NULL;
+}
+
+static json_t *pack_array(scanner_t *s, va_list *ap)
+{
+    json_t *array = json_array();
+    next_token(s);
+
+    while(s->token != ']') {
+        json_t *value;
+
+        if(!s->token) {
+            set_error(s, "<format>", "Unexpected end of format string");
+            goto error;
+        }
+
+        value = pack(s, ap);
+        if(!value)
+            goto error;
+
+        if(json_array_append_new(array, value)) {
+            set_error(s, "<internal>", "Unable to append to array");
+            goto error;
+        }
+
+        next_token(s);
+    }
+    return array;
+
+error:
+    json_decref(array);
+    return NULL;
+}
+
+static json_t *pack(scanner_t *s, va_list *ap)
+{
+    switch(s->token) {
+        case '{':
+            return pack_object(s, ap);
+
+        case '[':
+            return pack_array(s, ap);
+
+        case 's': /* string */
+        {
+            const char *str = va_arg(*ap, const char *);
+            if(!str) {
+                set_error(s, "<args>", "NULL string argument");
+                return NULL;
+            }
+            if(!utf8_check_string(str, -1)) {
+                set_error(s, "<args>", "Invalid UTF-8 string");
+                return NULL;
+            }
+            return json_string_nocheck(str);
+        }
+
+        case 'n': /* null */
+            return json_null();
+
+        case 'b': /* boolean */
+            return va_arg(*ap, int) ? json_true() : json_false();
+
+        case 'i': /* integer from int */
+            return json_integer(va_arg(*ap, int));
+
+        case 'I': /* integer from json_int_t */
+            return json_integer(va_arg(*ap, json_int_t));
+
+        case 'f': /* real */
+            return json_real(va_arg(*ap, double));
+
+        case 'O': /* a json_t object; increments refcount */
+            return json_incref(va_arg(*ap, json_t *));
+
+        case 'o': /* a json_t object; doesn't increment refcount */
+            return va_arg(*ap, json_t *);
+
+        default:
+            set_error(s, "<format>", "Unexpected format character '%c'",
+                      s->token);
+            return NULL;
+    }
+}
+
+static int unpack(scanner_t *s, json_t *root, va_list *ap);
+
+static int unpack_object(scanner_t *s, json_t *root, va_list *ap)
+{
+    int ret = -1;
+    int strict = 0;
+
+    /* Use a set (emulated by a hashtable) to check that all object
+       keys are accessed. Checking that the correct number of keys
+       were accessed is not enough, as the same key can be unpacked
+       multiple times.
+    */
+    hashtable_t key_set;
+
+    if(hashtable_init(&key_set, jsonp_hash_str, jsonp_str_equal, NULL, NULL)) {
+        set_error(s, "<internal>", "Out of memory");
+        return -1;
+    }
+
+    if(!json_is_object(root)) {
+        set_error(s, "<validation>", "Expected object, got %s",
+                  type_name(root));
+        goto out;
+    }
+    next_token(s);
+
+    while(s->token != '}') {
+        const char *key;
+        json_t *value;
+
+        if(strict != 0) {
+            set_error(s, "<format>", "Expected '}' after '%c', got '%c'",
+                      (strict == 1 ? '!' : '*'), s->token);
+            goto out;
+        }
+
+        if(!s->token) {
+            set_error(s, "<format>", "Unexpected end of format string");
+            goto out;
+        }
+
+        if(s->token == '!' || s->token == '*') {
+            strict = (s->token == '!' ? 1 : -1);
+            next_token(s);
+            continue;
+        }
+
+        if(s->token != 's') {
+            set_error(s, "<format>", "Expected format 's', got '%c'", s->token);
+            goto out;
+        }
+
+        key = va_arg(*ap, const char *);
+        if(!key) {
+            set_error(s, "<args>", "NULL object key");
+            goto out;
+        }
+
+        next_token(s);
+
+        value = json_object_get(root, key);
+        if(!value) {
+            set_error(s, "<validation>", "Object item not found: %s", key);
+            goto out;
+        }
+
+        if(unpack(s, value, ap))
+            goto out;
+
+        hashtable_set(&key_set, (void *)key, NULL);
+        next_token(s);
+    }
+
+    if(strict == 0 && (s->flags & JSON_STRICT))
+        strict = 1;
+
+    if(strict == 1 && key_set.size != json_object_size(root)) {
+        long diff = (long)json_object_size(root) - (long)key_set.size;
+        set_error(s, "<validation>", "%li object item(s) left unpacked", diff);
+        goto out;
+    }
+
+    ret = 0;
+
+out:
+    hashtable_close(&key_set);
+    return ret;
+}
+
+static int unpack_array(scanner_t *s, json_t *root, va_list *ap)
+{
+    size_t i = 0;
+    int strict = 0;
+
+    if(!json_is_array(root)) {
+        set_error(s, "<validation>", "Expected array, got %s", type_name(root));
+        return -1;
+    }
+    next_token(s);
+
+    while(s->token != ']') {
+        json_t *value;
+
+        if(strict != 0) {
+            set_error(s, "<format>", "Expected ']' after '%c', got '%c'",
+                      (strict == 1 ? '!' : '*'),
+                      s->token);
+            return -1;
+        }
+
+        if(!s->token) {
+            set_error(s, "<format>", "Unexpected end of format string");
+            return -1;
+        }
+
+        if(s->token == '!' || s->token == '*') {
+            strict = (s->token == '!' ? 1 : -1);
+            next_token(s);
+            continue;
+        }
+
+        if(!strchr(unpack_value_starters, s->token)) {
+            set_error(s, "<format>", "Unexpected format character '%c'",
+                      s->token);
+            return -1;
+        }
+
+        value = json_array_get(root, i);
+        if(!value) {
+            set_error(s, "<validation>", "Array index %lu out of range",
+                      (unsigned long)i);
+            return -1;
+        }
+
+        if(unpack(s, value, ap))
+            return -1;
+
+        next_token(s);
+        i++;
+    }
+
+    if(strict == 0 && (s->flags & JSON_STRICT))
+        strict = 1;
+
+    if(strict == 1 && i != json_array_size(root)) {
+        long diff = (long)json_array_size(root) - (long)i;
+        set_error(s, "<validation>", "%li array item(s) left unpacked", diff);
+        return -1;
+    }
+
+    return 0;
+}
+
+static int unpack(scanner_t *s, json_t *root, va_list *ap)
+{
+    switch(s->token)
+    {
+        case '{':
+            return unpack_object(s, root, ap);
+
+        case '[':
+            return unpack_array(s, root, ap);
+
+        case 's':
+            if(!json_is_string(root)) {
+                set_error(s, "<validation>", "Expected string, got %s",
+                          type_name(root));
+                return -1;
+            }
+
+            if(!(s->flags & JSON_VALIDATE_ONLY)) {
+                const char **str;
+
+                str = va_arg(*ap, const char **);
+                if(!str) {
+                    set_error(s, "<args>", "NULL string argument");
+                    return -1;
+                }
+
+                *str = json_string_value(root);
+            }
+            return 0;
+
+        case 'i':
+            if(!json_is_integer(root)) {
+                set_error(s, "<validation>", "Expected integer, got %s",
+                          type_name(root));
+                return -1;
+            }
+
+            if(!(s->flags & JSON_VALIDATE_ONLY))
+                *va_arg(*ap, int*) = json_integer_value(root);
+
+            return 0;
+
+        case 'I':
+            if(!json_is_integer(root)) {
+                set_error(s, "<validation>", "Expected integer, got %s",
+                          type_name(root));
+                return -1;
+            }
+
+            if(!(s->flags & JSON_VALIDATE_ONLY))
+                *va_arg(*ap, json_int_t*) = json_integer_value(root);
+
+            return 0;
+
+        case 'b':
+            if(!json_is_boolean(root)) {
+                set_error(s, "<validation>", "Expected true or false, got %s",
+                          type_name(root));
+                return -1;
+            }
+
+            if(!(s->flags & JSON_VALIDATE_ONLY))
+                *va_arg(*ap, int*) = json_is_true(root);
+
+            return 0;
+
+        case 'f':
+            if(!json_is_real(root)) {
+                set_error(s, "<validation>", "Expected real, got %s",
+                          type_name(root));
+                return -1;
+            }
+
+            if(!(s->flags & JSON_VALIDATE_ONLY))
+                *va_arg(*ap, double*) = json_real_value(root);
+
+            return 0;
+
+        case 'F':
+            if(!json_is_number(root)) {
+                set_error(s, "<validation>", "Expected real or integer, got %s",
+                          type_name(root));
+                return -1;
+            }
+
+            if(!(s->flags & JSON_VALIDATE_ONLY))
+                *va_arg(*ap, double*) = json_number_value(root);
+
+            return 0;
+
+        case 'O':
+            if(!(s->flags & JSON_VALIDATE_ONLY))
+                json_incref(root);
+            /* Fall through */
+
+        case 'o':
+            if(!(s->flags & JSON_VALIDATE_ONLY))
+                *va_arg(*ap, json_t**) = root;
+
+            return 0;
+
+        case 'n':
+            /* Never assign, just validate */
+            if(!json_is_null(root)) {
+                set_error(s, "<validation>", "Expected null, got %s",
+                          type_name(root));
+                return -1;
+            }
+            return 0;
+
+        default:
+            set_error(s, "<format>", "Unexpected format character '%c'",
+                      s->token);
+            return -1;
+    }
+}
+
+json_t *json_vpack_ex(json_error_t *error, size_t flags,
+                      const char *fmt, va_list ap)
+{
+    scanner_t s;
+    va_list ap_copy;
+    json_t *value;
+
+    if(!fmt || !*fmt) {
+        jsonp_error_init(error, "<format>");
+        jsonp_error_set(error, -1, -1, 0, "NULL or empty format string");
+        return NULL;
+    }
+    jsonp_error_init(error, NULL);
+
+    scanner_init(&s, error, flags, fmt);
+    next_token(&s);
+
+    va_copy(ap_copy, ap);
+    value = pack(&s, &ap_copy);
+    va_end(ap_copy);
+
+    if(!value)
+        return NULL;
+
+    next_token(&s);
+    if(s.token) {
+        json_decref(value);
+        set_error(&s, "<format>", "Garbage after format string");
+        return NULL;
+    }
+
+    return value;
+}
+
+json_t *json_pack_ex(json_error_t *error, size_t flags, const char *fmt, ...)
+{
+    json_t *value;
+    va_list ap;
+
+    va_start(ap, fmt);
+    value = json_vpack_ex(error, flags, fmt, ap);
+    va_end(ap);
+
+    return value;
+}
+
+json_t *json_pack(const char *fmt, ...)
+{
+    json_t *value;
+    va_list ap;
+
+    va_start(ap, fmt);
+    value = json_vpack_ex(NULL, 0, fmt, ap);
+    va_end(ap);
+
+    return value;
+}
+
+int json_vunpack_ex(json_t *root, json_error_t *error, size_t flags,
+                    const char *fmt, va_list ap)
+{
+    scanner_t s;
+    va_list ap_copy;
+
+    if(!root) {
+        jsonp_error_init(error, "<root>");
+        jsonp_error_set(error, -1, -1, 0, "NULL root value");
+        return -1;
+    }
+
+    if(!fmt || !*fmt) {
+        jsonp_error_init(error, "<format>");
+        jsonp_error_set(error, -1, -1, 0, "NULL or empty format string");
+        return -1;
+    }
+    jsonp_error_init(error, NULL);
+
+    scanner_init(&s, error, flags, fmt);
+    next_token(&s);
+
+    va_copy(ap_copy, ap);
+    if(unpack(&s, root, &ap_copy)) {
+        va_end(ap_copy);
+        return -1;
+    }
+    va_end(ap_copy);
+
+    next_token(&s);
+    if(s.token) {
+        set_error(&s, "<format>", "Garbage after format string");
+        return -1;
+    }
+
+    return 0;
+}
+
+int json_unpack_ex(json_t *root, json_error_t *error, size_t flags, const char *fmt, ...)
+{
+    int ret;
+    va_list ap;
+
+    va_start(ap, fmt);
+    ret = json_vunpack_ex(root, error, flags, fmt, ap);
+    va_end(ap);
+
+    return ret;
+}
+
+int json_unpack(json_t *root, const char *fmt, ...)
+{
+    int ret;
+    va_list ap;
+
+    va_start(ap, fmt);
+    ret = json_vunpack_ex(root, NULL, 0, fmt, ap);
+    va_end(ap);
+
+    return ret;
+}

src/strbuffer.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -8,18 +8,19 @@
 #define _GNU_SOURCE
 #include <stdlib.h>
 #include <string.h>
+#include "jansson_private.h"
 #include "strbuffer.h"
-#include "util.h"
 
 #define STRBUFFER_MIN_SIZE  16
 #define STRBUFFER_FACTOR    2
+#define STRBUFFER_SIZE_MAX  ((size_t)-1)
 
 int strbuffer_init(strbuffer_t *strbuff)
 {
     strbuff->size = STRBUFFER_MIN_SIZE;
     strbuff->length = 0;
 
-    strbuff->value = malloc(strbuff->size);
+    strbuff->value = jsonp_malloc(strbuff->size);
     if(!strbuff->value)
         return -1;
 
@@ -30,7 +31,7 @@ int strbuffer_init(strbuffer_t *strbuff)
 
 void strbuffer_close(strbuffer_t *strbuff)
 {
-    free(strbuff->value);
+    jsonp_free(strbuff->value);
     strbuff->size = 0;
     strbuff->length = 0;
     strbuff->value = NULL;
@@ -64,16 +65,31 @@ int strbuffer_append_byte(strbuffer_t *strbuff, char byte)
     return strbuffer_append_bytes(strbuff, &byte, 1);
 }
 
-int strbuffer_append_bytes(strbuffer_t *strbuff, const char *data, int size)
+int strbuffer_append_bytes(strbuffer_t *strbuff, const char *data, size_t size)
 {
-    if(strbuff->length + size >= strbuff->size)
+    if(size >= strbuff->size - strbuff->length)
     {
-        strbuff->size = max(strbuff->size * STRBUFFER_FACTOR,
-                            strbuff->length + size + 1);
+        size_t new_size;
+        char *new_value;
 
-        strbuff->value = realloc(strbuff->value, strbuff->size);
-        if(!strbuff->value)
+        /* avoid integer overflow */
+        if (strbuff->size > STRBUFFER_SIZE_MAX / STRBUFFER_FACTOR
+            || size > STRBUFFER_SIZE_MAX - 1
+            || strbuff->length > STRBUFFER_SIZE_MAX - 1 - size)
             return -1;
+
+        new_size = max(strbuff->size * STRBUFFER_FACTOR,
+                       strbuff->length + size + 1);
+
+        new_value = jsonp_malloc(new_size);
+        if(!new_value)
+            return -1;
+
+        memcpy(new_value, strbuff->value, strbuff->length);
+
+        jsonp_free(strbuff->value);
+        strbuff->value = new_value;
+        strbuff->size = new_size;
     }
 
     memcpy(strbuff->value + strbuff->length, data, size);

src/strbuffer.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -10,8 +10,8 @@
 
 typedef struct {
     char *value;
-    int length;   /* bytes used */
-    int size;     /* bytes allocated */
+    size_t length;   /* bytes used */
+    size_t size;     /* bytes allocated */
 } strbuffer_t;
 
 int strbuffer_init(strbuffer_t *strbuff);
@@ -24,7 +24,7 @@ char *strbuffer_steal_value(strbuffer_t *strbuff);
 
 int strbuffer_append(strbuffer_t *strbuff, const char *string);
 int strbuffer_append_byte(strbuffer_t *strbuff, char byte);
-int strbuffer_append_bytes(strbuffer_t *strbuff, const char *data, int size);
+int strbuffer_append_bytes(strbuffer_t *strbuff, const char *data, size_t size);
 
 char strbuffer_pop(strbuffer_t *strbuff);
 

src/strconv.c

@@ -0,0 +1,108 @@
+#include <assert.h>
+#include <errno.h>
+#include <stdio.h>
+#include <string.h>
+#include "jansson_private.h"
+#include "strbuffer.h"
+
+#if JSON_HAVE_LOCALECONV
+#include <locale.h>
+
+/*
+  - This code assumes that the decimal separator is exactly one
+    character.
+
+  - If setlocale() is called by another thread between the call to
+    localeconv() and the call to sprintf() or strtod(), the result may
+    be wrong. setlocale() is not thread-safe and should not be used
+    this way. Multi-threaded programs should use uselocale() instead.
+*/
+
+static void to_locale(strbuffer_t *strbuffer)
+{
+    const char *point;
+    char *pos;
+
+    point = localeconv()->decimal_point;
+    if(*point == '.') {
+        /* No conversion needed */
+        return;
+    }
+
+    pos = strchr(strbuffer->value, '.');
+    if(pos)
+        *pos = *point;
+}
+
+static void from_locale(char *buffer)
+{
+    const char *point;
+    char *pos;
+
+    point = localeconv()->decimal_point;
+    if(*point == '.') {
+        /* No conversion needed */
+        return;
+    }
+
+    pos = strchr(buffer, *point);
+    if(pos)
+        *pos = '.';
+}
+#endif
+
+int jsonp_strtod(strbuffer_t *strbuffer, double *out)
+{
+    double value;
+    char *end;
+
+#if JSON_HAVE_LOCALECONV
+    to_locale(strbuffer);
+#endif
+
+    errno = 0;
+    value = strtod(strbuffer->value, &end);
+    assert(end == strbuffer->value + strbuffer->length);
+
+    if(errno == ERANGE && value != 0) {
+        /* Overflow */
+        return -1;
+    }
+
+    *out = value;
+    return 0;
+}
+
+int jsonp_dtostr(char *buffer, size_t size, double value)
+{
+    int ret;
+    size_t length;
+
+    ret = snprintf(buffer, size, "%.17g", value);
+    if(ret < 0)
+        return -1;
+
+    length = (size_t)ret;
+    if(length >= size)
+        return -1;
+
+#if JSON_HAVE_LOCALECONV
+    from_locale(buffer);
+#endif
+
+    /* Make sure there's a dot or 'e' in the output. Otherwise
+       a real is converted to an integer when decoding */
+    if(strchr(buffer, '.') == NULL &&
+       strchr(buffer, 'e') == NULL)
+    {
+        if(length + 2 >= size) {
+            /* No space to append ".0" */
+            return -1;
+        }
+        buffer[length] = '.';
+        buffer[length + 1] = '0';
+        length += 2;
+    }
+
+    return (int)length;
+}

src/utf.c

@@ -1,12 +1,12 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
  */
 
 #include <string.h>
-#include <stdint.h>
+#include "utf.h"
 
 int utf8_encode(int32_t codepoint, char *buffer, int *size)
 {

src/utf.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -8,6 +8,26 @@
 #ifndef UTF_H
 #define UTF_H
 
+#ifdef HAVE_CONFIG_H
+#include <config.h>
+
+#ifdef HAVE_INTTYPES_H
+/* inttypes.h includes stdint.h in a standard environment, so there's
+no need to include stdint.h separately. If inttypes.h doesn't define
+int32_t, it's defined in config.h. */
+#include <inttypes.h>
+#endif /* HAVE_INTTYPES_H */
+
+#else /* !HAVE_CONFIG_H */
+#ifdef _WIN32
+typedef int int32_t;
+#else /* !_WIN32 */
+/* Assume a standard environment */
+#include <inttypes.h>
+#endif /* _WIN32 */
+
+#endif /* HAVE_CONFIG_H */
+
 int utf8_encode(int codepoint, char *buffer, int *size);
 
 int utf8_check_first(char byte);

src/util.h

@@ -1,13 +0,0 @@
-/*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
- *
- * Jansson is free software; you can redistribute it and/or modify
- * it under the terms of the MIT license. See LICENSE for details.
- */
-
-#ifndef UTIL_H
-#define UTIL_H
-
-#define max(a, b)  ((a) > (b) ? (a) : (b))
-
-#endif

src/value.c

@@ -1,11 +1,13 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
  */
 
 #define _GNU_SOURCE
+
+#include <stddef.h>
 #include <stdlib.h>
 #include <string.h>
 
@@ -13,10 +15,9 @@
 #include "hashtable.h"
 #include "jansson_private.h"
 #include "utf.h"
-#include "util.h"
 
 
-static inline void json_init(json_t *json, json_type type)
+static JSON_INLINE void json_init(json_t *json, json_type type)
 {
     json->type = type;
     json->refcount = 1;
@@ -25,13 +26,15 @@ static inline void json_init(json_t *json, json_type type)
 
 /*** object ***/
 
-static unsigned int hash_string(const void *key)
+/* From http://www.cse.yorku.ca/~oz/hash.html */
+size_t jsonp_hash_str(const void *ptr)
 {
-    const char *str = (const char *)key;
-    unsigned int hash = 5381;
-    unsigned int c;
+    const char *str = (const char *)ptr;
 
-    while((c = (unsigned int)*str))
+    size_t hash = 5381;
+    size_t c;
+
+    while((c = (size_t)*str))
     {
         hash = ((hash << 5) + hash) + c;
         str++;
@@ -40,9 +43,26 @@ static unsigned int hash_string(const void *key)
     return hash;
 }
 
-static int string_equal(const void *key1, const void *key2)
+int jsonp_str_equal(const void *ptr1, const void *ptr2)
 {
-    return strcmp((const char *)key1, (const char *)key2) == 0;
+    return strcmp((const char *)ptr1, (const char *)ptr2) == 0;
+}
+
+/* This macro just returns a pointer that's a few bytes backwards from
+   string. This makes it possible to pass a pointer to object_key_t
+   when only the string inside it is used, without actually creating
+   an object_key_t instance. */
+#define string_to_key(string)  container_of(string, object_key_t, key)
+
+static size_t hash_key(const void *ptr)
+{
+    return jsonp_hash_str(((const object_key_t *)ptr)->key);
+}
+
+static int key_equal(const void *ptr1, const void *ptr2)
+{
+    return jsonp_str_equal(((const object_key_t *)ptr1)->key,
+                           ((const object_key_t *)ptr2)->key);
 }
 
 static void value_decref(void *value)
@@ -52,18 +72,20 @@ static void value_decref(void *value)
 
 json_t *json_object(void)
 {
-    json_object_t *object = malloc(sizeof(json_object_t));
+    json_object_t *object = jsonp_malloc(sizeof(json_object_t));
     if(!object)
         return NULL;
     json_init(&object->json, JSON_OBJECT);
 
-    if(hashtable_init(&object->hashtable, hash_string, string_equal,
-                      free, value_decref))
+    if(hashtable_init(&object->hashtable,
+                      hash_key, key_equal,
+                      jsonp_free, value_decref))
     {
-        free(object);
+        jsonp_free(object);
         return NULL;
     }
 
+    object->serial = 0;
     object->visited = 0;
 
     return &object->json;
@@ -72,15 +94,15 @@ json_t *json_object(void)
 static void json_delete_object(json_object_t *object)
 {
     hashtable_close(&object->hashtable);
-    free(object);
+    jsonp_free(object);
 }
 
-unsigned int json_object_size(const json_t *json)
+size_t json_object_size(const json_t *json)
 {
     json_object_t *object;
 
     if(!json_is_object(json))
-        return -1;
+        return 0;
 
     object = json_to_object(json);
     return object->hashtable.size;
@@ -94,12 +116,13 @@ json_t *json_object_get(const json_t *json, const char *key)
         return NULL;
 
     object = json_to_object(json);
-    return hashtable_get(&object->hashtable, key);
+    return hashtable_get(&object->hashtable, string_to_key(key));
 }
 
 int json_object_set_new_nocheck(json_t *json, const char *key, json_t *value)
 {
     json_object_t *object;
+    object_key_t *k;
 
     if(!key || !value)
         return -1;
@@ -111,7 +134,20 @@ int json_object_set_new_nocheck(json_t *json, const char *key, json_t *value)
     }
     object = json_to_object(json);
 
-    if(hashtable_set(&object->hashtable, strdup(key), value))
+    /* offsetof(...) returns the size of object_key_t without the
+       last, flexible member. This way, the correct amount is
+       allocated. */
+    k = jsonp_malloc(offsetof(object_key_t, key) + strlen(key) + 1);
+    if(!k)
+    {
+        json_decref(value);
+        return -1;
+    }
+
+    k->serial = object->serial++;
+    strcpy(k->key, key);
+
+    if(hashtable_set(&object->hashtable, k, value))
     {
         json_decref(value);
         return -1;
@@ -139,7 +175,7 @@ int json_object_del(json_t *json, const char *key)
         return -1;
 
     object = json_to_object(json);
-    return hashtable_del(&object->hashtable, key);
+    return hashtable_del(&object->hashtable, string_to_key(key));
 }
 
 int json_object_clear(json_t *json)
@@ -190,6 +226,17 @@ void *json_object_iter(json_t *json)
     return hashtable_iter(&object->hashtable);
 }
 
+void *json_object_iter_at(json_t *json, const char *key)
+{
+    json_object_t *object;
+
+    if(!key || !json_is_object(json))
+        return NULL;
+
+    object = json_to_object(json);
+    return hashtable_iter_at(&object->hashtable, string_to_key(key));
+}
+
 void *json_object_iter_next(json_t *json, void *iter)
 {
     json_object_t *object;
@@ -201,12 +248,20 @@ void *json_object_iter_next(json_t *json, void *iter)
     return hashtable_iter_next(&object->hashtable, iter);
 }
 
+const object_key_t *jsonp_object_iter_fullkey(void *iter)
+{
+    if(!iter)
+        return NULL;
+
+    return hashtable_iter_key(iter);
+}
+
 const char *json_object_iter_key(void *iter)
 {
     if(!iter)
         return NULL;
 
-    return (const char *)hashtable_iter_key(iter);
+    return jsonp_object_iter_fullkey(iter)->key;
 }
 
 json_t *json_object_iter_value(void *iter)
@@ -217,6 +272,19 @@ json_t *json_object_iter_value(void *iter)
     return (json_t *)hashtable_iter_value(iter);
 }
 
+int json_object_iter_set_new(json_t *json, void *iter, json_t *value)
+{
+    json_object_t *object;
+
+    if(!json_is_object(json) || !iter || !value)
+        return -1;
+
+    object = json_to_object(json);
+    hashtable_iter_set(&object->hashtable, iter, value);
+
+    return 0;
+}
+
 static int json_object_equal(json_t *object1, json_t *object2)
 {
     void *iter;
@@ -298,7 +366,7 @@ static json_t *json_object_deep_copy(json_t *object)
 
 json_t *json_array(void)
 {
-    json_array_t *array = malloc(sizeof(json_array_t));
+    json_array_t *array = jsonp_malloc(sizeof(json_array_t));
     if(!array)
         return NULL;
     json_init(&array->json, JSON_ARRAY);
@@ -306,9 +374,9 @@ json_t *json_array(void)
     array->entries = 0;
     array->size = 8;
 
-    array->table = malloc(array->size * sizeof(json_t *));
+    array->table = jsonp_malloc(array->size * sizeof(json_t *));
     if(!array->table) {
-        free(array);
+        jsonp_free(array);
         return NULL;
     }
 
@@ -319,16 +387,16 @@ json_t *json_array(void)
 
 static void json_delete_array(json_array_t *array)
 {
-    unsigned int i;
+    size_t i;
 
     for(i = 0; i < array->entries; i++)
         json_decref(array->table[i]);
 
-    free(array->table);
-    free(array);
+    jsonp_free(array->table);
+    jsonp_free(array);
 }
 
-unsigned int json_array_size(const json_t *json)
+size_t json_array_size(const json_t *json)
 {
     if(!json_is_array(json))
         return 0;
@@ -336,7 +404,7 @@ unsigned int json_array_size(const json_t *json)
     return json_to_array(json)->entries;
 }
 
-json_t *json_array_get(const json_t *json, unsigned int index)
+json_t *json_array_get(const json_t *json, size_t index)
 {
     json_array_t *array;
     if(!json_is_array(json))
@@ -349,7 +417,7 @@ json_t *json_array_get(const json_t *json, unsigned int index)
     return array->table[index];
 }
 
-int json_array_set_new(json_t *json, unsigned int index, json_t *value)
+int json_array_set_new(json_t *json, size_t index, json_t *value)
 {
     json_array_t *array;
 
@@ -375,24 +443,24 @@ int json_array_set_new(json_t *json, unsigned int index, json_t *value)
     return 0;
 }
 
-static void array_move(json_array_t *array, unsigned int dest,
-                       unsigned int src, unsigned int count)
+static void array_move(json_array_t *array, size_t dest,
+                       size_t src, size_t count)
 {
     memmove(&array->table[dest], &array->table[src], count * sizeof(json_t *));
 }
 
-static void array_copy(json_t **dest, unsigned int dpos,
-                       json_t **src, unsigned int spos,
-                       unsigned int count)
+static void array_copy(json_t **dest, size_t dpos,
+                       json_t **src, size_t spos,
+                       size_t count)
 {
     memcpy(&dest[dpos], &src[spos], count * sizeof(json_t *));
 }
 
 static json_t **json_array_grow(json_array_t *array,
-                                unsigned int amount,
+                                size_t amount,
                                 int copy)
 {
-    unsigned int new_size;
+    size_t new_size;
     json_t **old_table, **new_table;
 
     if(array->entries + amount <= array->size)
@@ -401,7 +469,7 @@ static json_t **json_array_grow(json_array_t *array,
     old_table = array->table;
 
     new_size = max(array->size + amount, array->size * 2);
-    new_table = malloc(new_size * sizeof(json_t *));
+    new_table = jsonp_malloc(new_size * sizeof(json_t *));
     if(!new_table)
         return NULL;
 
@@ -410,7 +478,7 @@ static json_t **json_array_grow(json_array_t *array,
 
     if(copy) {
         array_copy(array->table, 0, old_table, 0, array->entries);
-        free(old_table);
+        jsonp_free(old_table);
         return array->table;
     }
 
@@ -442,7 +510,7 @@ int json_array_append_new(json_t *json, json_t *value)
     return 0;
 }
 
-int json_array_insert_new(json_t *json, unsigned int index, json_t *value)
+int json_array_insert_new(json_t *json, size_t index, json_t *value)
 {
     json_array_t *array;
     json_t **old_table;
@@ -471,7 +539,7 @@ int json_array_insert_new(json_t *json, unsigned int index, json_t *value)
         array_copy(array->table, 0, old_table, 0, index);
         array_copy(array->table, index + 1, old_table, index,
                    array->entries - index);
-        free(old_table);
+        jsonp_free(old_table);
     }
     else
         array_move(array, index + 1, index, array->entries - index);
@@ -482,7 +550,7 @@ int json_array_insert_new(json_t *json, unsigned int index, json_t *value)
     return 0;
 }
 
-int json_array_remove(json_t *json, unsigned int index)
+int json_array_remove(json_t *json, size_t index)
 {
     json_array_t *array;
 
@@ -504,7 +572,7 @@ int json_array_remove(json_t *json, unsigned int index)
 int json_array_clear(json_t *json)
 {
     json_array_t *array;
-    unsigned int i;
+    size_t i;
 
     if(!json_is_array(json))
         return -1;
@@ -520,7 +588,7 @@ int json_array_clear(json_t *json)
 int json_array_extend(json_t *json, json_t *other_json)
 {
     json_array_t *array, *other;
-    unsigned int i;
+    size_t i;
 
     if(!json_is_array(json) || !json_is_array(other_json))
         return -1;
@@ -541,7 +609,7 @@ int json_array_extend(json_t *json, json_t *other_json)
 
 static int json_array_equal(json_t *array1, json_t *array2)
 {
-    unsigned int i, size;
+    size_t i, size;
 
     size = json_array_size(array1);
     if(size != json_array_size(array2))
@@ -564,7 +632,7 @@ static int json_array_equal(json_t *array1, json_t *array2)
 static json_t *json_array_copy(json_t *array)
 {
     json_t *result;
-    unsigned int i;
+    size_t i;
 
     result = json_array();
     if(!result)
@@ -579,7 +647,7 @@ static json_t *json_array_copy(json_t *array)
 static json_t *json_array_deep_copy(json_t *array)
 {
     json_t *result;
-    unsigned int i;
+    size_t i;
 
     result = json_array();
     if(!result)
@@ -600,14 +668,14 @@ json_t *json_string_nocheck(const char *value)
     if(!value)
         return NULL;
 
-    string = malloc(sizeof(json_string_t));
+    string = jsonp_malloc(sizeof(json_string_t));
     if(!string)
         return NULL;
     json_init(&string->json, JSON_STRING);
 
-    string->value = strdup(value);
+    string->value = jsonp_strdup(value);
     if(!string->value) {
-        free(string);
+        jsonp_free(string);
         return NULL;
     }
 
@@ -635,12 +703,15 @@ int json_string_set_nocheck(json_t *json, const char *value)
     char *dup;
     json_string_t *string;
 
-    dup = strdup(value);
+    if(!json_is_string(json) || !value)
+        return -1;
+
+    dup = jsonp_strdup(value);
     if(!dup)
         return -1;
 
     string = json_to_string(json);
-    free(string->value);
+    jsonp_free(string->value);
     string->value = dup;
 
     return 0;
@@ -656,8 +727,8 @@ int json_string_set(json_t *json, const char *value)
 
 static void json_delete_string(json_string_t *string)
 {
-    free(string->value);
-    free(string);
+    jsonp_free(string->value);
+    jsonp_free(string);
 }
 
 static int json_string_equal(json_t *string1, json_t *string2)
@@ -673,9 +744,9 @@ static json_t *json_string_copy(json_t *string)
 
 /*** integer ***/
 
-json_t *json_integer(int value)
+json_t *json_integer(json_int_t value)
 {
-    json_integer_t *integer = malloc(sizeof(json_integer_t));
+    json_integer_t *integer = jsonp_malloc(sizeof(json_integer_t));
     if(!integer)
         return NULL;
     json_init(&integer->json, JSON_INTEGER);
@@ -684,7 +755,7 @@ json_t *json_integer(int value)
     return &integer->json;
 }
 
-int json_integer_value(const json_t *json)
+json_int_t json_integer_value(const json_t *json)
 {
     if(!json_is_integer(json))
         return 0;
@@ -692,7 +763,7 @@ int json_integer_value(const json_t *json)
     return json_to_integer(json)->value;
 }
 
-int json_integer_set(json_t *json, int value)
+int json_integer_set(json_t *json, json_int_t value)
 {
     if(!json_is_integer(json))
         return -1;
@@ -704,7 +775,7 @@ int json_integer_set(json_t *json, int value)
 
 static void json_delete_integer(json_integer_t *integer)
 {
-    free(integer);
+    jsonp_free(integer);
 }
 
 static int json_integer_equal(json_t *integer1, json_t *integer2)
@@ -722,7 +793,7 @@ static json_t *json_integer_copy(json_t *integer)
 
 json_t *json_real(double value)
 {
-    json_real_t *real = malloc(sizeof(json_real_t));
+    json_real_t *real = jsonp_malloc(sizeof(json_real_t));
     if(!real)
         return NULL;
     json_init(&real->json, JSON_REAL);
@@ -751,7 +822,7 @@ int json_real_set(json_t *json, double value)
 
 static void json_delete_real(json_real_t *real)
 {
-    free(real);
+    jsonp_free(real);
 }
 
 static int json_real_equal(json_t *real1, json_t *real2)
@@ -782,30 +853,21 @@ double json_number_value(const json_t *json)
 
 json_t *json_true(void)
 {
-    static json_t the_true = {
-        .type = JSON_TRUE,
-        .refcount = (unsigned int)1
-    };
+    static json_t the_true = {JSON_TRUE, (size_t)-1};
     return &the_true;
 }
 
 
 json_t *json_false(void)
 {
-    static json_t the_false = {
-        .type = JSON_FALSE,
-        .refcount = (unsigned int)1
-    };
+    static json_t the_false = {JSON_FALSE, (size_t)-1};
     return &the_false;
 }
 
 
 json_t *json_null(void)
 {
-    static json_t the_null = {
-        .type = JSON_NULL,
-        .refcount = (unsigned int)1
-    };
+    static json_t the_null = {JSON_NULL, (size_t)-1};
     return &the_null;
 }
 

test/.gitignore

@@ -1,9 +1,16 @@
 logs
 bin/json_process
 suites/api/test_array
-suites/api/test_equal
 suites/api/test_copy
+suites/api/test_cpp
+suites/api/test_dump
+suites/api/test_dump_callback
+suites/api/test_equal
 suites/api/test_load
+suites/api/test_loadb
+suites/api/test_memory_funcs
 suites/api/test_number
 suites/api/test_object
+suites/api/test_pack
 suites/api/test_simple
+suites/api/test_unpack

test/bin/json_process.c

@@ -1,14 +1,24 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include <config.h>
+#endif
+
 #include <stdio.h>
 #include <stdlib.h>
+#include <string.h>
+#include <ctype.h>
 #include <jansson.h>
 
+#if HAVE_LOCALE_H
+#include <locale.h>
+#endif
+
 static int getenv_int(const char *name)
 {
     char *value, *end;
@@ -25,14 +35,38 @@ static int getenv_int(const char *name)
     return (int)result;
 }
 
+/* Return a pointer to the first non-whitespace character of str.
+   Modifies str so that all trailing whitespace characters are
+   replaced by '\0'. */
+static const char *strip(char *str)
+{
+    size_t length;
+    char *result = str;
+    while(*result && isspace(*result))
+        result++;
+
+    length = strlen(result);
+    if(length == 0)
+        return result;
+
+    while(isspace(result[length - 1]))
+        result[--length] = '\0';
+
+    return result;
+}
+
 int main(int argc, char *argv[])
 {
     int indent = 0;
-    unsigned int flags = 0;
+    size_t flags = 0;
 
     json_t *json;
     json_error_t error;
 
+#if HAVE_SETLOCALE
+    setlocale(LC_ALL, "");
+#endif
+
     if(argc != 1) {
         fprintf(stderr, "usage: %s\n", argv[0]);
         return 2;
@@ -53,12 +87,45 @@ int main(int argc, char *argv[])
     if(getenv_int("JSON_ENSURE_ASCII"))
         flags |= JSON_ENSURE_ASCII;
 
+    if(getenv_int("JSON_PRESERVE_ORDER"))
+        flags |= JSON_PRESERVE_ORDER;
+
     if(getenv_int("JSON_SORT_KEYS"))
         flags |= JSON_SORT_KEYS;
 
-    json = json_loadf(stdin, &error);
+    if(getenv_int("STRIP")) {
+        /* Load to memory, strip leading and trailing whitespace */
+        size_t size = 0, used = 0;
+        char *buffer = NULL;
+
+        while(1) {
+            int count;
+
+            size = (size == 0 ? 128 : size * 2);
+            buffer = realloc(buffer, size);
+            if(!buffer) {
+                fprintf(stderr, "Unable to allocate %d bytes\n", (int)size);
+                return 1;
+            }
+
+            count = fread(buffer + used, 1, size - used, stdin);
+            if(count < size - used) {
+                buffer[used + count] = '\0';
+                break;
+            }
+            used += count;
+        }
+
+        json = json_loads(strip(buffer), 0, &error);
+        free(buffer);
+    }
+    else
+        json = json_loadf(stdin, 0, &error);
+
     if(!json) {
-        fprintf(stderr, "%d\n%s\n", error.line, error.text);
+        fprintf(stderr, "%d %d %d\n%s\n",
+                error.line, error.column, error.position,
+                error.text);
         return 1;
     }
 

test/run-suites

@@ -20,6 +20,8 @@ if [ -z "$SUITES" ]; then
     done
 fi
 
+[ -z "$STOP" ] && STOP=0
+
 export suites_srcdir=$top_srcdir/test/suites
 export suites_builddir=suites
 export scriptdir=$top_srcdir/test/scripts
@@ -34,6 +36,7 @@ for suite in $SUITES; do
         passed=$(($passed+1))
     else
         failed=$(($failed+1))
+        [ $STOP -eq 1 ] && break
     fi
 done
 

test/scripts/run-tests.sh

@@ -1,4 +1,4 @@
-# Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+# Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
 #
 # Jansson is free software; you can redistribute it and/or modify
 # it under the terms of the MIT license. See LICENSE for details.
@@ -12,6 +12,7 @@ suite_log=$logdir/$suite_name
 
 
 [ -z "$VERBOSE" ] && VERBOSE=0
+[ -z "$STOP" ] && STOP=0
 
 . $scriptdir/valgrind.sh
 
@@ -29,29 +30,46 @@ for test_path in $suite_srcdir/*; do
     rm -rf $test_log
     mkdir -p $test_log
     if [ $VERBOSE -eq 1 ]; then
-        echo -n "$name... "
+        printf '%s... ' "$test_name"
     fi
 
-    if run_test; then
-        # Success
-        if [ $VERBOSE -eq 1 ]; then
-            echo "ok"
-        else
-            echo -n "."
-        fi
-        rm -rf $test_log
-    else
-        # Failure
-        if [ $VERBOSE -eq 1 ]; then
-            echo "FAILED"
-        else
-            echo -n "F"
-        fi
-    fi
+    run_test
+    case $? in
+        0)
+            # Success
+            if [ $VERBOSE -eq 1 ]; then
+                printf 'ok\n'
+            else
+                printf '.'
+            fi
+            rm -rf $test_log
+            ;;
+
+        77)
+            # Skip
+            if [ $VERBOSE -eq 1 ]; then
+                printf 'skipped\n'
+            else
+                printf 'S'
+            fi
+            rm -rf $test_log
+            ;;
+
+        *)
+            # Failure
+            if [ $VERBOSE -eq 1 ]; then
+                printf 'FAILED\n'
+            else
+                printf 'F'
+            fi
+
+            [ $STOP -eq 1 ] && break
+            ;;
+    esac
 done
 
 if [ $VERBOSE -eq 0 ]; then
-    echo
+    printf '\n'
 fi
 
 if [ -n "$(ls -A $suite_log)" ]; then

test/scripts/valgrind.sh

@@ -1,4 +1,4 @@
-# Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+# Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
 #
 # Jansson is free software; you can redistribute it and/or modify
 # it under the terms of the MIT license. See LICENSE for details.

test/suites/Makefile.am

@@ -1,2 +1,2 @@
 SUBDIRS = api
-EXTRA_DIST = invalid invalid-strip invalid-unicode valid valid-strip
+EXTRA_DIST = invalid invalid-unicode valid

test/suites/api/Makefile.am

@@ -2,19 +2,31 @@ EXTRA_DIST = run
 
 check_PROGRAMS = \
 	test_array \
-	test_equal \
 	test_copy \
+	test_dump \
+	test_dump_callback \
+	test_equal \
 	test_load \
-	test_simple \
+	test_loadb \
+	test_memory_funcs \
 	test_number \
-	test_object
+	test_object \
+	test_pack \
+	test_simple \
+	test_unpack
 
 test_array_SOURCES = test_array.c util.h
 test_copy_SOURCES = test_copy.c util.h
+test_dump_SOURCES = test_dump.c util.h
+test_dump_callback_SOURCES = test_dump_callback.c util.h
 test_load_SOURCES = test_load.c util.h
-test_simple_SOURCES = test_simple.c util.h
+test_loadb_SOURCES = test_loadb.c util.h
+test_memory_funcs_SOURCES = test_memory_funcs.c util.h
 test_number_SOURCES = test_number.c util.h
 test_object_SOURCES = test_object.c util.h
+test_pack_SOURCES = test_pack.c util.h
+test_simple_SOURCES = test_simple.c util.h
+test_unpack_SOURCES = test_unpack.c util.h
 
 AM_CPPFLAGS = -I$(top_srcdir)/src
 AM_CFLAGS = -Wall -Werror

test/suites/api/check-exports

@@ -39,18 +39,29 @@ json_object_del
 json_object_clear
 json_object_update
 json_object_iter
+json_object_iter_at
 json_object_iter_next
 json_object_iter_key
 json_object_iter_value
+json_object_iter_set_new
 json_dumps
 json_dumpf
 json_dump_file
+json_dump_callback
 json_loads
 json_loadf
 json_load_file
+json_loadb
 json_equal
 json_copy
 json_deep_copy
+json_pack
+json_pack_ex
+json_vpack_ex
+json_unpack
+json_unpack_ex
+json_vunpack_ex
+json_set_alloc_funcs
 EOF
 
 # The list of functions are not exported in the library because they
@@ -80,7 +91,10 @@ EOF
 
 SOFILE="../src/.libs/libjansson.so"
 
-nm -D $SOFILE | grep ' T ' | cut -d' ' -f3 | sort >$test_log/output
+nm -D $SOFILE >/dev/null >$test_log/symbols 2>/dev/null \
+    || exit 77  # Skip if "nm -D" doesn't seem to work
+
+grep ' T ' $test_log/symbols | cut -d' ' -f3 | sort >$test_log/output
 
 if ! cmp -s $test_log/exports $test_log/output; then
     diff -u $test_log/exports $test_log/output >&2

test/suites/api/run

@@ -1,23 +1,29 @@
 #!/bin/sh
 #
-# Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+# Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
 #
 # Jansson is free software; you can redistribute it and/or modify
 # it under the terms of the MIT license. See LICENSE for details.
 
 is_test() {
-    [ "${test_name%.c}" != "$test_name" ] && return 0
-    [ -x $test_path -a ! -f $test_path.c ] && return 0
-    return 1
+    case "$test_name" in
+        *.c|check-exports)
+            return 0
+            ;;
+        *)
+            return 1
+            ;;
+    esac
 }
 
 run_test() {
-    if [ -x $test_path ]; then
+    if [ "$test_name" = "check-exports" ]; then
         test_log=$test_log $test_path >$test_log/stdout 2>$test_log/stderr
     else
         $test_runner $suite_builddir/${test_name%.c} \
             >$test_log/stdout \
-            2>$test_log/stderr
+            2>$test_log/stderr \
+            || return 1
         valgrind_check $test_log/stderr || return 1
     fi
 }

test/suites/api/test_array.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -387,7 +387,7 @@ static void test_circular()
 }
 
 
-int main()
+static void run_tests()
 {
     test_misc();
     test_insert();
@@ -395,6 +395,4 @@ int main()
     test_clear();
     test_extend();
     test_circular();
-
-    return 0;
 }

test/suites/api/test_copy.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -174,9 +174,9 @@ static void test_copy_array(void)
     const char *json_array_text = "[1, \"foo\", 3.141592, {\"foo\": \"bar\"}]";
 
     json_t *array, *copy;
-    unsigned int i;
+    size_t i;
 
-    array = json_loads(json_array_text, NULL);
+    array = json_loads(json_array_text, 0, NULL);
     if(!array)
         fail("unable to parse an array");
 
@@ -203,9 +203,9 @@ static void test_deep_copy_array(void)
     const char *json_array_text = "[1, \"foo\", 3.141592, {\"foo\": \"bar\"}]";
 
     json_t *array, *copy;
-    unsigned int i;
+    size_t i;
 
-    array = json_loads(json_array_text, NULL);
+    array = json_loads(json_array_text, 0, NULL);
     if(!array)
         fail("unable to parse an array");
 
@@ -235,7 +235,7 @@ static void test_copy_object(void)
     json_t *object, *copy;
     void *iter;
 
-    object = json_loads(json_object_text, NULL);
+    object = json_loads(json_object_text, 0, NULL);
     if(!object)
         fail("unable to parse an object");
 
@@ -275,7 +275,7 @@ static void test_deep_copy_object(void)
     json_t *object, *copy;
     void *iter;
 
-    object = json_loads(json_object_text, NULL);
+    object = json_loads(json_object_text, 0, NULL);
     if(!object)
         fail("unable to parse an object");
 
@@ -307,7 +307,7 @@ static void test_deep_copy_object(void)
     json_decref(copy);
 }
 
-int main()
+static void run_tests()
 {
     test_copy_simple();
     test_deep_copy_simple();
@@ -315,5 +315,4 @@ int main()
     test_deep_copy_array();
     test_copy_object();
     test_deep_copy_object();
-    return 0;
 }

test/suites/api/test_dump.c

@@ -0,0 +1,141 @@
+/*
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
+ *
+ * Jansson is free software; you can redistribute it and/or modify
+ * it under the terms of the MIT license. See LICENSE for details.
+ */
+
+#include <jansson.h>
+#include <string.h>
+#include "util.h"
+
+static void encode_twice()
+{
+    /* Encode an empty object/array, add an item, encode again */
+
+    json_t *json;
+    char *result;
+
+    json = json_object();
+    result = json_dumps(json, 0);
+    if(!result || strcmp(result, "{}"))
+      fail("json_dumps failed");
+    free(result);
+
+    json_object_set_new(json, "foo", json_integer(5));
+    result = json_dumps(json, 0);
+    if(!result || strcmp(result, "{\"foo\": 5}"))
+      fail("json_dumps failed");
+    free(result);
+
+    json_decref(json);
+
+    json = json_array();
+    result = json_dumps(json, 0);
+    if(!result || strcmp(result, "[]"))
+      fail("json_dumps failed");
+    free(result);
+
+    json_array_append_new(json, json_integer(5));
+    result = json_dumps(json, 0);
+    if(!result || strcmp(result, "[5]"))
+      fail("json_dumps failed");
+    free(result);
+
+    json_decref(json);
+}
+
+static void circular_references()
+{
+    /* Construct a JSON object/array with a circular reference:
+
+       object: {"a": {"b": {"c": <circular reference to $.a>}}}
+       array: [[[<circular reference to the $[0] array>]]]
+
+       Encode it, remove the circular reference and encode again.
+    */
+
+    json_t *json;
+    char *result;
+
+    json = json_object();
+    json_object_set_new(json, "a", json_object());
+    json_object_set_new(json_object_get(json, "a"), "b", json_object());
+    json_object_set(json_object_get(json_object_get(json, "a"), "b"), "c",
+                    json_object_get(json, "a"));
+
+    if(json_dumps(json, 0))
+        fail("json_dumps encoded a circular reference!");
+
+    json_object_del(json_object_get(json_object_get(json, "a"), "b"), "c");
+
+    result = json_dumps(json, 0);
+    if(!result || strcmp(result, "{\"a\": {\"b\": {}}}"))
+        fail("json_dumps failed!");
+    free(result);
+
+    json_decref(json);
+
+    json = json_array();
+    json_array_append_new(json, json_array());
+    json_array_append_new(json_array_get(json, 0), json_array());
+    json_array_append(json_array_get(json_array_get(json, 0), 0),
+                      json_array_get(json, 0));
+
+    if(json_dumps(json, 0))
+        fail("json_dumps encoded a circular reference!");
+
+    json_array_remove(json_array_get(json_array_get(json, 0), 0), 0);
+
+    result = json_dumps(json, 0);
+    if(!result || strcmp(result, "[[[]]]"))
+        fail("json_dumps failed!");
+    free(result);
+
+    json_decref(json);
+}
+
+static void encode_other_than_array_or_object()
+{
+    /* Encoding anything other than array or object should only
+     * succeed if the JSON_ENCODE_ANY flag is used */
+
+    json_t *json;
+    FILE *fp = NULL;
+    char *result;
+
+    json = json_string("foo");
+    if(json_dumps(json, 0) != NULL)
+        fail("json_dumps encoded a string!");
+    if(json_dumpf(json, fp, 0) == 0)
+        fail("json_dumpf encoded a string!");
+
+    result = json_dumps(json, JSON_ENCODE_ANY);
+    if(!result || strcmp(result, "\"foo\"") != 0)
+        fail("json_dumps failed to encode a string with JSON_ENCODE_ANY");
+
+    free(result);
+    json_decref(json);
+
+    json = json_integer(42);
+    if(json_dumps(json, 0) != NULL)
+        fail("json_dumps encoded an integer!");
+    if(json_dumpf(json, fp, 0) == 0)
+        fail("json_dumpf encoded an integer!");
+
+    result = json_dumps(json, JSON_ENCODE_ANY);
+    if(!result || strcmp(result, "42") != 0)
+        fail("json_dumps failed to encode an integer with JSON_ENCODE_ANY");
+
+    free(result);
+    json_decref(json);
+
+
+}
+
+static void run_tests()
+{
+    encode_twice();
+    circular_references();
+    encode_other_than_array_or_object();
+}

test/suites/api/test_dump_callback.c

@@ -0,0 +1,81 @@
+/*
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
+ *
+ * Jansson is free software; you can redistribute it and/or modify
+ * it under the terms of the MIT license. See LICENSE for details.
+ */
+
+#include <jansson.h>
+#include <string.h>
+#include <stdlib.h>
+#include "util.h"
+
+struct my_sink {
+    char *buf;
+    size_t off;
+    size_t cap;
+};
+
+static int my_writer(const char *buffer, size_t len, void *data) {
+    struct my_sink *s = data;
+    if (len > s->cap - s->off) {
+        return -1;
+    }
+    memcpy(s->buf + s->off, buffer, len);
+    s->off += len;
+    return 0;
+}
+
+static void run_tests()
+{
+    struct my_sink s;
+    json_t *json;
+    const char str[] = "[\"A\", {\"B\": \"C\", \"e\": false}, 1, null, \"foo\"]";
+    char *dumped_to_string;
+
+    json = json_loads(str, 0, NULL);
+    if(!json) {
+        fail("json_loads failed");
+    }
+
+    dumped_to_string = json_dumps(json, 0);
+    if (!dumped_to_string) {
+        json_decref(json);
+        fail("json_dumps failed");
+    }
+
+    s.off = 0;
+    s.cap = strlen(dumped_to_string);
+    s.buf = malloc(s.cap);
+    if (!s.buf) {
+        json_decref(json);
+        free(dumped_to_string);
+        fail("malloc failed");
+    }
+
+    if (json_dump_callback(json, my_writer, &s, 0) == -1) {
+        json_decref(json);
+        free(dumped_to_string);
+        free(s.buf);
+        fail("json_dump_callback failed on an exact-length sink buffer");
+    }
+
+    if (strncmp(dumped_to_string, s.buf, s.off) != 0) {
+        json_decref(json);
+        free(dumped_to_string);
+        free(s.buf);
+        fail("json_dump_callback and json_dumps did not produce identical output");
+    }
+
+    s.off = 1;
+    if (json_dump_callback(json, my_writer, &s, 0) != -1) {
+        json_decref(json);
+        free(dumped_to_string);
+        free(s.buf);
+        fail("json_dump_callback succeeded on a short buffer when it should have failed");
+    }
+
+    json_decref(json);
+    free(dumped_to_string);
+    free(s.buf);
+}

test/suites/api/test_equal.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -167,8 +167,8 @@ static void test_equal_complex()
 "    \"array\": [\"foo\", false, null, 1.234]"
 "}";
 
-    value1 = json_loads(complex_json, NULL);
-    value2 = json_loads(complex_json, NULL);
+    value1 = json_loads(complex_json, 0, NULL);
+    value2 = json_loads(complex_json, 0, NULL);
     if(!value1 || !value2)
         fail("unable to parse JSON");
     if(!json_equal(value1, value2))
@@ -180,11 +180,10 @@ static void test_equal_complex()
     /* TODO: There's no negative test case here */
 }
 
-int main()
+static void run_tests()
 {
     test_equal_simple();
     test_equal_array();
     test_equal_object();
     test_equal_complex();
-    return 0;
 }

test/suites/api/test_load.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -9,16 +9,50 @@
 #include <string.h>
 #include "util.h"
 
-int main()
+static void file_not_found()
 {
     json_t *json;
     json_error_t error;
 
-    json = json_load_file("/path/to/nonexistent/file.json", &error);
+    json = json_load_file("/path/to/nonexistent/file.json", 0, &error);
+    if(json)
+        fail("json_load_file returned non-NULL for a nonexistent file");
     if(error.line != -1)
         fail("json_load_file returned an invalid line number");
     if(strcmp(error.text, "unable to open /path/to/nonexistent/file.json: No such file or directory") != 0)
         fail("json_load_file returned an invalid error message");
-
-    return 0;
+}
+
+static void reject_duplicates()
+{
+    json_error_t error;
+
+    if(json_loads("{\"foo\": 1, \"foo\": 2}", JSON_REJECT_DUPLICATES, &error))
+        fail("json_loads did not detect a duplicate key");
+    check_error("duplicate object key near '\"foo\"'", "<string>", 1, 16, 16);
+}
+
+static void disable_eof_check()
+{
+    json_error_t error;
+    json_t *json;
+
+    const char *text = "{\"foo\": 1} garbage";
+
+    if(json_loads(text, 0, &error))
+        fail("json_loads did not detect garbage after JSON text");
+    check_error("end of file expected near 'garbage'", "<string>", 1, 18, 18);
+
+    json = json_loads(text, JSON_DISABLE_EOF_CHECK, &error);
+    if(!json)
+        fail("json_loads failed with JSON_DISABLE_EOF_CHECK");
+
+    json_decref(json);
+}
+
+static void run_tests()
+{
+    file_not_found();
+    reject_duplicates();
+    disable_eof_check();
 }

test/suites/api/test_loadb.c

@@ -0,0 +1,36 @@
+/*
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
+ *
+ * Jansson is free software; you can redistribute it and/or modify
+ * it under the terms of the MIT license. See LICENSE for details.
+ */
+
+#include <jansson.h>
+#include <string.h>
+#include "util.h"
+
+static void run_tests()
+{
+    json_t *json;
+    json_error_t error;
+    const char str[] = "[\"A\", {\"B\": \"C\"}, 1, 2, 3]garbage";
+    size_t len = strlen(str) - strlen("garbage");
+
+    json = json_loadb(str, len, 0, &error);
+    if(!json) {
+        fail("json_loadb failed on a valid JSON buffer");
+    }
+    json_decref(json);
+
+    json = json_loadb(str, len - 1, 0, &error);
+    if (json) {
+        json_decref(json);
+        fail("json_loadb should have failed on an incomplete buffer, but it didn't");
+    }
+    if(error.line != 1) {
+        fail("json_loadb returned an invalid line number on fail");
+    }
+    if(strcmp(error.text, "']' expected near end of file") != 0) {
+        fail("json_loadb returned an invalid error message for an unclosed top-level array");
+    }
+}

test/suites/api/test_memory_funcs.c

@@ -0,0 +1,82 @@
+#include <string.h>
+#include <jansson.h>
+
+#include "util.h"
+
+static int malloc_called = 0;
+static int free_called = 0;
+
+/* helper */
+static void create_and_free_complex_object()
+{
+    json_t *obj;
+
+    obj = json_pack("{s:i,s:n,s:b,s:b,s:{s:s},s:[i,i,i]",
+                    "foo", 42,
+                    "bar",
+                    "baz", 1,
+                    "qux", 0,
+                    "alice", "bar", "baz",
+                    "bob", 9, 8, 7);
+
+    json_decref(obj);
+}
+
+static void *my_malloc(size_t size)
+{
+    malloc_called += 1;
+    return malloc(size);
+}
+
+static void my_free(void *ptr)
+{
+    free_called += 1;
+    free(ptr);
+}
+
+static void test_simple()
+{
+    json_set_alloc_funcs(my_malloc, my_free);
+    create_and_free_complex_object();
+
+    if(malloc_called != 27 || free_called != 27)
+        fail("Custom allocation failed");
+}
+
+
+/*
+  Test the secure memory functions code given in the API reference
+  documentation, but by using plain memset instead of
+  guaranteed_memset().
+*/
+
+static void *secure_malloc(size_t size)
+{
+    /* Store the memory area size in the beginning of the block */
+    void *ptr = malloc(size + 8);
+    *((size_t *)ptr) = size;
+    return ptr + 8;
+}
+
+static void secure_free(void *ptr)
+{
+    size_t size;
+
+    ptr -= 8;
+    size = *((size_t *)ptr);
+
+    /*guaranteed_*/memset(ptr, 0, size);
+    free(ptr);
+}
+
+static void test_secure_funcs(void)
+{
+    json_set_alloc_funcs(secure_malloc, secure_free);
+    create_and_free_complex_object();
+}
+
+static void run_tests()
+{
+    test_simple();
+    test_secure_funcs();
+}

test/suites/api/test_number.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -8,7 +8,7 @@
 #include <jansson.h>
 #include "util.h"
 
-int main()
+static void run_tests()
 {
     json_t *integer, *real;
     int i;
@@ -39,6 +39,4 @@ int main()
 
     json_decref(integer);
     json_decref(real);
-
-    return 0;
 }

test/suites/api/test_object.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -205,10 +205,98 @@ static void test_set_nocheck()
     json_decref(object);
 }
 
+static void test_iterators()
+{
+    json_t *object, *foo, *bar, *baz;
+    void *iter;
+
+    if(json_object_iter(NULL))
+        fail("able to iterate over NULL");
+
+    if(json_object_iter_next(NULL, NULL))
+        fail("able to increment an iterator on a NULL object");
+
+    object = json_object();
+    foo = json_string("foo");
+    bar = json_string("bar");
+    baz = json_string("baz");
+    if(!object || !foo || !bar || !bar)
+        fail("unable to create values");
+
+    if(json_object_iter_next(object, NULL))
+        fail("able to increment a NULL iterator");
+
+    if(json_object_set(object, "a", foo) ||
+       json_object_set(object, "b", bar) ||
+       json_object_set(object, "c", baz))
+        fail("unable to populate object");
+
+    iter = json_object_iter(object);
+    if(!iter)
+        fail("unable to get iterator");
+    if(strcmp(json_object_iter_key(iter), "a"))
+        fail("iterating failed: wrong key");
+    if(json_object_iter_value(iter) != foo)
+        fail("iterating failed: wrong value");
+
+    iter = json_object_iter_next(object, iter);
+    if(!iter)
+        fail("unable to increment iterator");
+    if(strcmp(json_object_iter_key(iter), "b"))
+        fail("iterating failed: wrong key");
+    if(json_object_iter_value(iter) != bar)
+        fail("iterating failed: wrong value");
+
+    iter = json_object_iter_next(object, iter);
+    if(!iter)
+        fail("unable to increment iterator");
+    if(strcmp(json_object_iter_key(iter), "c"))
+        fail("iterating failed: wrong key");
+    if(json_object_iter_value(iter) != baz)
+        fail("iterating failed: wrong value");
+
+    if(json_object_iter_next(object, iter) != NULL)
+        fail("able to iterate over the end");
+
+    if(json_object_iter_at(object, "foo"))
+        fail("json_object_iter_at() succeeds for non-existent key");
+
+    iter = json_object_iter_at(object, "b");
+    if(!iter)
+        fail("json_object_iter_at() fails for an existing key");
+
+    if(strcmp(json_object_iter_key(iter), "b"))
+        fail("iterating failed: wrong key");
+    if(json_object_iter_value(iter) != bar)
+        fail("iterating failed: wrong value");
+
+    iter = json_object_iter_next(object, iter);
+    if(!iter)
+        fail("unable to increment iterator");
+    if(strcmp(json_object_iter_key(iter), "c"))
+        fail("iterating failed: wrong key");
+    if(json_object_iter_value(iter) != baz)
+        fail("iterating failed: wrong value");
+
+    if(json_object_iter_set(object, iter, bar))
+        fail("unable to set value at iterator");
+
+    if(strcmp(json_object_iter_key(iter), "c"))
+        fail("json_object_iter_key() fails after json_object_iter_set()");
+    if(json_object_iter_value(iter) != bar)
+        fail("json_object_iter_value() fails after json_object_iter_set()");
+    if(json_object_get(object, "c") != bar)
+        fail("json_object_get() fails after json_object_iter_set()");
+
+    json_decref(object);
+    json_decref(foo);
+    json_decref(bar);
+    json_decref(baz);
+}
+
 static void test_misc()
 {
     json_t *object, *string, *other_string, *value;
-    void *iter;
 
     object = json_object();
     string = json_string("test");
@@ -231,17 +319,6 @@ static void test_misc()
     if(!json_object_set(object, "a", NULL))
         fail("able to set NULL value");
 
-    iter = json_object_iter(object);
-    if(!iter)
-        fail("unable to get iterator");
-
-    if(strcmp(json_object_iter_key(iter), "a"))
-        fail("iterating failed: wrong key");
-    if(json_object_iter_value(iter) != string)
-        fail("iterating failed: wrong value");
-    if(json_object_iter_next(object, iter) != NULL)
-        fail("able to iterate over the end");
-
     /* invalid UTF-8 in key */
     if(!json_object_set(object, "a\xefz", string))
         fail("able to set invalid unicode key");
@@ -325,13 +402,48 @@ static void test_misc()
     json_decref(object);
 }
 
-int main()
+static void test_preserve_order()
+{
+    json_t *object;
+    char *result;
+
+    const char *expected = "{\"foobar\": 1, \"bazquux\": 6, \"lorem ipsum\": 3, \"sit amet\": 5, \"helicopter\": 7}";
+
+    object = json_object();
+
+    json_object_set_new(object, "foobar", json_integer(1));
+    json_object_set_new(object, "bazquux", json_integer(2));
+    json_object_set_new(object, "lorem ipsum", json_integer(3));
+    json_object_set_new(object, "dolor", json_integer(4));
+    json_object_set_new(object, "sit amet", json_integer(5));
+
+    /* changing a value should preserve the order */
+    json_object_set_new(object, "bazquux", json_integer(6));
+
+    /* deletion shouldn't change the order of others */
+    json_object_del(object, "dolor");
+
+    /* add a new item just to make sure */
+    json_object_set_new(object, "helicopter", json_integer(7));
+
+    result = json_dumps(object, JSON_PRESERVE_ORDER);
+
+    if(strcmp(expected, result) != 0) {
+        fprintf(stderr, "%s != %s", expected, result);
+        fail("JSON_PRESERVE_ORDER doesn't work");
+    }
+
+    free(result);
+    json_decref(object);
+}
+
+static void run_tests()
 {
     test_misc();
     test_clear();
     test_update();
     test_circular();
     test_set_nocheck();
-
-    return 0;
+    test_iterators();
+    test_preserve_order();
 }

test/suites/api/test_pack.c

@@ -0,0 +1,230 @@
+/*
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2010-2011 Graeme Smecher <graeme.smecher@mail.mcgill.ca>
+ *
+ * Jansson is free software; you can redistribute it and/or modify
+ * it under the terms of the MIT license. See LICENSE for details.
+ */
+
+#include <string.h>
+#include <jansson.h>
+#include <stdio.h>
+#include "util.h"
+
+static void run_tests()
+{
+    json_t *value;
+    int i;
+    json_error_t error;
+
+    /*
+     * Simple, valid json_pack cases
+     */
+
+    /* true */
+    value = json_pack("b", 1);
+    if(!json_is_true(value))
+        fail("json_pack boolean failed");
+    if(value->refcount != (ssize_t)-1)
+        fail("json_pack boolean refcount failed");
+    json_decref(value);
+
+    /* false */
+    value = json_pack("b", 0);
+    if(!json_is_false(value))
+        fail("json_pack boolean failed");
+    if(value->refcount != (ssize_t)-1)
+        fail("json_pack boolean refcount failed");
+    json_decref(value);
+
+    /* null */
+    value = json_pack("n");
+    if(!json_is_null(value))
+        fail("json_pack null failed");
+    if(value->refcount != (ssize_t)-1)
+        fail("json_pack null refcount failed");
+    json_decref(value);
+
+    /* integer */
+    value = json_pack("i", 1);
+    if(!json_is_integer(value) || json_integer_value(value) != 1)
+        fail("json_pack integer failed");
+    if(value->refcount != (ssize_t)1)
+        fail("json_pack integer refcount failed");
+    json_decref(value);
+
+    /* integer from json_int_t */
+    value = json_pack("I", (json_int_t)555555);
+    if(!json_is_integer(value) || json_integer_value(value) != 555555)
+        fail("json_pack json_int_t failed");
+    if(value->refcount != (ssize_t)1)
+        fail("json_pack integer refcount failed");
+    json_decref(value);
+
+    /* real */
+    value = json_pack("f", 1.0);
+    if(!json_is_real(value) || json_real_value(value) != 1.0)
+        fail("json_pack real failed");
+    if(value->refcount != (ssize_t)1)
+        fail("json_pack real refcount failed");
+    json_decref(value);
+
+    /* string */
+    value = json_pack("s", "test");
+    if(!json_is_string(value) || strcmp("test", json_string_value(value)))
+        fail("json_pack string failed");
+    if(value->refcount != (ssize_t)1)
+        fail("json_pack string refcount failed");
+    json_decref(value);
+
+    /* empty object */
+    value = json_pack("{}", 1.0);
+    if(!json_is_object(value) || json_object_size(value) != 0)
+        fail("json_pack empty object failed");
+    if(value->refcount != (ssize_t)1)
+        fail("json_pack empty object refcount failed");
+    json_decref(value);
+
+    /* empty list */
+    value = json_pack("[]", 1.0);
+    if(!json_is_array(value) || json_array_size(value) != 0)
+        fail("json_pack empty list failed");
+    if(value->refcount != (ssize_t)1)
+        fail("json_pack empty list failed");
+    json_decref(value);
+
+    /* non-incref'd object */
+    value = json_pack("o", json_integer(1));
+    if(!json_is_integer(value) || json_integer_value(value) != 1)
+        fail("json_pack object failed");
+    if(value->refcount != (ssize_t)1)
+        fail("json_pack integer refcount failed");
+    json_decref(value);
+
+    /* incref'd object */
+    value = json_pack("O", json_integer(1));
+    if(!json_is_integer(value) || json_integer_value(value) != 1)
+        fail("json_pack object failed");
+    if(value->refcount != (ssize_t)2)
+        fail("json_pack integer refcount failed");
+    json_decref(value);
+    json_decref(value);
+
+    /* simple object */
+    value = json_pack("{s:[]}", "foo");
+    if(!json_is_object(value) || json_object_size(value) != 1)
+        fail("json_pack array failed");
+    if(!json_is_array(json_object_get(value, "foo")))
+        fail("json_pack array failed");
+    if(json_object_get(value, "foo")->refcount != (ssize_t)1)
+        fail("json_pack object refcount failed");
+    json_decref(value);
+
+    /* simple array */
+    value = json_pack("[i,i,i]", 0, 1, 2);
+    if(!json_is_array(value) || json_array_size(value) != 3)
+        fail("json_pack object failed");
+    for(i=0; i<3; i++)
+    {
+        if(!json_is_integer(json_array_get(value, i)) ||
+           json_integer_value(json_array_get(value, i)) != i)
+
+            fail("json_pack integer array failed");
+    }
+    json_decref(value);
+
+    /* Whitespace; regular string */
+    value = json_pack(" s ", "test");
+    if(!json_is_string(value) || strcmp("test", json_string_value(value)))
+        fail("json_pack string (with whitespace) failed");
+    json_decref(value);
+
+    /* Whitespace; empty array */
+    value = json_pack("[ ]");
+    if(!json_is_array(value) || json_array_size(value) != 0)
+        fail("json_pack empty array (with whitespace) failed");
+    json_decref(value);
+
+    /* Whitespace; array */
+    value = json_pack("[ i , i,  i ] ", 1, 2, 3);
+    if(!json_is_array(value) || json_array_size(value) != 3)
+        fail("json_pack array (with whitespace) failed");
+    json_decref(value);
+
+    /*
+     * Invalid cases
+     */
+
+    /* newline in format string */
+    if(json_pack_ex(&error, 0, "{\n\n1"))
+        fail("json_pack failed to catch invalid format '1'");
+    check_error("Expected format 's', got '1'", "<format>", 3, 1, 4);
+
+    /* mismatched open/close array/object */
+    if(json_pack_ex(&error, 0, "[}"))
+        fail("json_pack failed to catch mismatched '}'");
+    check_error("Unexpected format character '}'", "<format>", 1, 2, 2);
+
+    if(json_pack_ex(&error, 0, "{]"))
+        fail("json_pack failed to catch mismatched ']'");
+    check_error("Expected format 's', got ']'", "<format>", 1, 2, 2);
+
+    /* missing close array */
+    if(json_pack_ex(&error, 0, "["))
+        fail("json_pack failed to catch missing ']'");
+    check_error("Unexpected end of format string", "<format>", 1, 2, 2);
+
+    /* missing close object */
+    if(json_pack_ex(&error, 0, "{"))
+        fail("json_pack failed to catch missing '}'");
+    check_error("Unexpected end of format string", "<format>", 1, 2, 2);
+
+    /* garbage after format string */
+    if(json_pack_ex(&error, 0, "[i]a", 42))
+        fail("json_pack failed to catch garbage after format string");
+    check_error("Garbage after format string", "<format>", 1, 4, 4);
+
+    if(json_pack_ex(&error, 0, "ia", 42))
+        fail("json_pack failed to catch garbage after format string");
+    check_error("Garbage after format string", "<format>", 1, 2, 2);
+
+    /* NULL string */
+    if(json_pack_ex(&error, 0, "s", NULL))
+        fail("json_pack failed to catch null argument string");
+    check_error("NULL string argument", "<args>", 1, 1, 1);
+
+    /* NULL format */
+    if(json_pack_ex(&error, 0, NULL))
+        fail("json_pack failed to catch NULL format string");
+    check_error("NULL or empty format string", "<format>", -1, -1, 0);
+
+    /* NULL key */
+    if(json_pack_ex(&error, 0, "{s:i}", NULL, 1))
+        fail("json_pack failed to catch NULL key");
+    check_error("NULL object key", "<args>", 1, 2, 2);
+
+    /* More complicated checks for row/columns */
+    if(json_pack_ex(&error, 0, "{ {}: s }", "foo"))
+        fail("json_pack failed to catch object as key");
+    check_error("Expected format 's', got '{'", "<format>", 1, 3, 3);
+
+    /* Complex object */
+    if(json_pack_ex(&error, 0, "{ s: {},  s:[ii{} }", "foo", "bar", 12, 13))
+        fail("json_pack failed to catch missing ]");
+    check_error("Unexpected format character '}'", "<format>", 1, 19, 19);
+
+    /* Complex array */
+    if(json_pack_ex(&error, 0, "[[[[[   [[[[[  [[[[ }]]]] ]]]] ]]]]]"))
+        fail("json_pack failed to catch extra }");
+    check_error("Unexpected format character '}'", "<format>", 1, 21, 21);
+
+    /* Invalid UTF-8 in object key */
+    if(json_pack_ex(&error, 0, "{s:i}", "\xff\xff", 42))
+        fail("json_pack failed to catch invalid UTF-8 in an object key");
+    check_error("Invalid UTF-8 in object key", "<args>", 1, 2, 2);
+
+    /* Invalid UTF-8 in a string */
+    if(json_pack_ex(&error, 0, "{s:s}", "foo", "\xff\xff"))
+        fail("json_pack failed to catch invalid UTF-8 in a string");
+    check_error("Invalid UTF-8 string", "<args>", 1, 4, 4);
+}

test/suites/api/test_simple.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
@@ -10,7 +10,7 @@
 #include "util.h"
 
 /* Call the simple functions not covered by other tests of the public API */
-int main()
+static void run_tests()
 {
     json_t *value;
 
@@ -150,5 +150,34 @@ int main()
         fail("json_null failed");
     json_decref(value);
 
-    return 0;
+    /* Test reference counting on singletons (true, false, null) */
+    value = json_true();
+    if(value->refcount != (size_t)-1)
+      fail("refcounting true works incorrectly");
+    json_decref(value);
+    if(value->refcount != (size_t)-1)
+      fail("refcounting true works incorrectly");
+    json_incref(value);
+    if(value->refcount != (size_t)-1)
+      fail("refcounting true works incorrectly");
+
+    value = json_false();
+    if(value->refcount != (size_t)-1)
+      fail("refcounting false works incorrectly");
+    json_decref(value);
+    if(value->refcount != (size_t)-1)
+      fail("refcounting false works incorrectly");
+    json_incref(value);
+    if(value->refcount != (size_t)-1)
+      fail("refcounting false works incorrectly");
+
+    value = json_null();
+    if(value->refcount != (size_t)-1)
+      fail("refcounting null works incorrectly");
+    json_decref(value);
+    if(value->refcount != (size_t)-1)
+      fail("refcounting null works incorrectly");
+    json_incref(value);
+    if(value->refcount != (size_t)-1)
+      fail("refcounting null works incorrectly");
 }

test/suites/api/test_unpack.c

@@ -0,0 +1,339 @@
+/*
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2010-2011 Graeme Smecher <graeme.smecher@mail.mcgill.ca>
+ *
+ * Jansson is free software; you can redistribute it and/or modify
+ * it under the terms of the MIT license. See LICENSE for details.
+ */
+
+#include <string.h>
+#include <jansson.h>
+#include <stdio.h>
+#include "util.h"
+
+static void run_tests()
+{
+    json_t *j, *j2;
+    int i1, i2, i3;
+    json_int_t I1;
+    int rv;
+    double f;
+    char *s;
+
+    json_error_t error;
+
+    /*
+     * Simple, valid json_pack cases
+     */
+
+    /* true */
+    rv = json_unpack(json_true(), "b", &i1);
+    if(rv || !i1)
+        fail("json_unpack boolean failed");
+
+    /* false */
+    rv = json_unpack(json_false(), "b", &i1);
+    if(rv || i1)
+        fail("json_unpack boolean failed");
+
+    /* null */
+    if(json_unpack(json_null(), "n"))
+        fail("json_unpack null failed");
+
+    /* integer */
+    j = json_integer(42);
+    rv = json_unpack(j, "i", &i1);
+    if(rv || i1 != 42)
+        fail("json_unpack integer failed");
+    json_decref(j);
+
+    /* json_int_t */
+    j = json_integer(5555555);
+    rv = json_unpack(j, "I", &I1);
+    if(rv || I1 != 5555555)
+        fail("json_unpack json_int_t failed");
+    json_decref(j);
+
+    /* real */
+    j = json_real(1.7);
+    rv = json_unpack(j, "f", &f);
+    if(rv || f != 1.7)
+        fail("json_unpack real failed");
+    json_decref(j);
+
+    /* number */
+    j = json_integer(12345);
+    rv = json_unpack(j, "F", &f);
+    if(rv || f != 12345.0)
+        fail("json_unpack (real or) integer failed");
+    json_decref(j);
+
+    j = json_real(1.7);
+    rv = json_unpack(j, "F", &f);
+    if(rv || f != 1.7)
+        fail("json_unpack real (or integer) failed");
+    json_decref(j);
+
+    /* string */
+    j = json_string("foo");
+    rv = json_unpack(j, "s", &s);
+    if(rv || strcmp(s, "foo"))
+        fail("json_unpack string failed");
+    json_decref(j);
+
+    /* empty object */
+    j = json_object();
+    if(json_unpack(j, "{}"))
+        fail("json_unpack empty object failed");
+    json_decref(j);
+
+    /* empty list */
+    j = json_array();
+    if(json_unpack(j, "[]"))
+        fail("json_unpack empty list failed");
+    json_decref(j);
+
+    /* non-incref'd object */
+    j = json_object();
+    rv = json_unpack(j, "o", &j2);
+    if(j2 != j || j->refcount != 1)
+        fail("json_unpack object failed");
+    json_decref(j);
+
+    /* incref'd object */
+    j = json_object();
+    rv = json_unpack(j, "O", &j2);
+    if(j2 != j || j->refcount != 2)
+        fail("json_unpack object failed");
+    json_decref(j);
+    json_decref(j);
+
+    /* simple object */
+    j = json_pack("{s:i}", "foo", 42);
+    rv = json_unpack(j, "{s:i}", "foo", &i1);
+    if(rv || i1 != 42)
+        fail("json_unpack simple object failed");
+    json_decref(j);
+
+    /* simple array */
+    j = json_pack("[iii]", 1, 2, 3);
+    rv = json_unpack(j, "[i,i,i]", &i1, &i2, &i3);
+    if(rv || i1 != 1 || i2 != 2 || i3 != 3)
+        fail("json_unpack simple array failed");
+    json_decref(j);
+
+    /* object with many items & strict checking */
+    j = json_pack("{s:i, s:i, s:i}", "a", 1, "b", 2, "c", 3);
+    rv = json_unpack(j, "{s:i, s:i, s:i}", "a", &i1, "b", &i2, "c", &i3);
+    if(rv || i1 != 1 || i2 != 2 || i3 != 3)
+        fail("json_unpack object with many items failed");
+    json_decref(j);
+
+    /*
+     * Invalid cases
+     */
+
+    j = json_integer(42);
+    if(!json_unpack_ex(j, &error, 0, "z"))
+        fail("json_unpack succeeded with invalid format character");
+    check_error("Unexpected format character 'z'", "<format>", 1, 1, 1);
+
+    if(!json_unpack_ex(NULL, &error, 0, "[i]"))
+        fail("json_unpack succeeded with NULL root");
+    check_error("NULL root value", "<root>", -1, -1, 0);
+    json_decref(j);
+
+    /* mismatched open/close array/object */
+    j = json_pack("[]");
+    if(!json_unpack_ex(j, &error, 0, "[}"))
+        fail("json_unpack failed to catch mismatched ']'");
+    check_error("Unexpected format character '}'", "<format>", 1, 2, 2);
+    json_decref(j);
+
+    j = json_pack("{}");
+    if(!json_unpack_ex(j, &error, 0, "{]"))
+        fail("json_unpack failed to catch mismatched '}'");
+    check_error("Expected format 's', got ']'", "<format>", 1, 2, 2);
+    json_decref(j);
+
+    /* missing close array */
+    j = json_pack("[]");
+    if(!json_unpack_ex(j, &error, 0, "["))
+        fail("json_unpack failed to catch missing ']'");
+    check_error("Unexpected end of format string", "<format>", 1, 2, 2);
+    json_decref(j);
+
+    /* missing close object */
+    j = json_pack("{}");
+    if(!json_unpack_ex(j, &error, 0, "{"))
+        fail("json_unpack failed to catch missing '}'");
+    check_error("Unexpected end of format string", "<format>", 1, 2, 2);
+    json_decref(j);
+
+    /* garbage after format string */
+    j = json_pack("[i]", 42);
+    if(!json_unpack_ex(j, &error, 0, "[i]a", &i1))
+        fail("json_unpack failed to catch garbage after format string");
+    check_error("Garbage after format string", "<format>", 1, 4, 4);
+    json_decref(j);
+
+    j = json_integer(12345);
+    if(!json_unpack_ex(j, &error, 0, "ia", &i1))
+        fail("json_unpack failed to catch garbage after format string");
+    check_error("Garbage after format string", "<format>", 1, 2, 2);
+    json_decref(j);
+
+    /* NULL format string */
+    j = json_pack("[]");
+    if(!json_unpack_ex(j, &error, 0, NULL))
+        fail("json_unpack failed to catch null format string");
+    check_error("NULL or empty format string", "<format>", -1, -1, 0);
+    json_decref(j);
+
+    /* NULL string pointer */
+    j = json_string("foobie");
+    if(!json_unpack_ex(j, &error, 0, "s", NULL))
+        fail("json_unpack failed to catch null string pointer");
+    check_error("NULL string argument", "<args>", 1, 1, 1);
+    json_decref(j);
+
+    /* invalid types */
+    j = json_integer(42);
+    j2 = json_string("foo");
+    if(!json_unpack_ex(j, &error, 0, "s"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected string, got integer", "<validation>", 1, 1, 1);
+
+    if(!json_unpack_ex(j, &error, 0, "n"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected null, got integer", "<validation>", 1, 1, 1);
+
+    if(!json_unpack_ex(j, &error, 0, "b"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected true or false, got integer", "<validation>", 1, 1, 1);
+
+    if(!json_unpack_ex(j2, &error, 0, "i"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected integer, got string", "<validation>", 1, 1, 1);
+
+    if(!json_unpack_ex(j2, &error, 0, "I"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected integer, got string", "<validation>", 1, 1, 1);
+
+    if(!json_unpack_ex(j, &error, 0, "f"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected real, got integer", "<validation>", 1, 1, 1);
+
+    if(!json_unpack_ex(j2, &error, 0, "F"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected real or integer, got string", "<validation>", 1, 1, 1);
+
+    if(!json_unpack_ex(j, &error, 0, "[i]"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected array, got integer", "<validation>", 1, 1, 1);
+
+    if(!json_unpack_ex(j, &error, 0, "{si}", "foo"))
+        fail("json_unpack failed to catch invalid type");
+    check_error("Expected object, got integer", "<validation>", 1, 1, 1);
+
+    json_decref(j);
+    json_decref(j2);
+
+    /* Array index out of range */
+    j = json_pack("[i]", 1);
+    if(!json_unpack_ex(j, &error, 0, "[ii]", &i1, &i2))
+        fail("json_unpack failed to catch index out of array bounds");
+    check_error("Array index 1 out of range", "<validation>", 1, 3, 3);
+    json_decref(j);
+
+    /* NULL object key */
+    j = json_pack("{si}", "foo", 42);
+    if(!json_unpack_ex(j, &error, 0, "{si}", NULL, &i1))
+        fail("json_unpack failed to catch null string pointer");
+    check_error("NULL object key", "<args>", 1, 2, 2);
+    json_decref(j);
+
+    /* Object key not found */
+    j = json_pack("{si}", "foo", 42);
+    if(!json_unpack_ex(j, &error, 0, "{si}", "baz", &i1))
+        fail("json_unpack failed to catch null string pointer");
+    check_error("Object item not found: baz", "<validation>", 1, 3, 3);
+    json_decref(j);
+
+    /*
+     * Strict validation
+     */
+
+    j = json_pack("[iii]", 1, 2, 3);
+    rv = json_unpack(j, "[iii!]", &i1, &i2, &i3);
+    if(rv || i1 != 1 || i2 != 2 || i3 != 3)
+        fail("json_unpack array with strict validation failed");
+    json_decref(j);
+
+    j = json_pack("[iii]", 1, 2, 3);
+    if(!json_unpack_ex(j, &error, 0, "[ii!]", &i1, &i2))
+        fail("json_unpack array with strict validation failed");
+    check_error("1 array item(s) left unpacked", "<validation>", 1, 5, 5);
+    json_decref(j);
+
+    /* Like above, but with JSON_STRICT instead of '!' format */
+    j = json_pack("[iii]", 1, 2, 3);
+    if(!json_unpack_ex(j, &error, JSON_STRICT, "[ii]", &i1, &i2))
+        fail("json_unpack array with strict validation failed");
+    check_error("1 array item(s) left unpacked", "<validation>", 1, 4, 4);
+    json_decref(j);
+
+    j = json_pack("{s:s, s:i}", "foo", "bar", "baz", 42);
+    rv = json_unpack(j, "{sssi!}", "foo", &s, "baz", &i1);
+    if(rv || strcmp(s, "bar") != 0 || i1 != 42)
+        fail("json_unpack object with strict validation failed");
+    json_decref(j);
+
+    /* Unpack the same item twice */
+    j = json_pack("{s:s, s:i}", "foo", "bar", "baz", 42);
+    if(!json_unpack_ex(j, &error, 0, "{s:s,s:s!}", "foo", &s, "foo", &s))
+        fail("json_unpack object with strict validation failed");
+    check_error("1 object item(s) left unpacked", "<validation>", 1, 10, 10);
+    json_decref(j);
+
+    j = json_pack("[i,{s:i,s:n},[i,i]]", 1, "foo", 2, "bar", 3, 4);
+    if(json_unpack_ex(j, NULL, JSON_STRICT | JSON_VALIDATE_ONLY,
+                      "[i{sisn}[ii]]", "foo", "bar"))
+        fail("json_unpack complex value with strict validation failed");
+    json_decref(j);
+
+    /* ! and * must be last */
+    j = json_pack("[ii]", 1, 2);
+    if(!json_unpack_ex(j, &error, 0, "[i!i]", &i1, &i2))
+        fail("json_unpack failed to catch ! in the middle of an array");
+    check_error("Expected ']' after '!', got 'i'", "<format>", 1, 4, 4);
+
+    if(!json_unpack_ex(j, &error, 0, "[i*i]", &i1, &i2))
+        fail("json_unpack failed to catch * in the middle of an array");
+    check_error("Expected ']' after '*', got 'i'", "<format>", 1, 4, 4);
+    json_decref(j);
+
+    j = json_pack("{sssi}", "foo", "bar", "baz", 42);
+    if(!json_unpack_ex(j, &error, 0, "{ss!si}", "foo", &s, "baz", &i1))
+        fail("json_unpack failed to catch ! in the middle of an object");
+    check_error("Expected '}' after '!', got 's'", "<format>", 1, 5, 5);
+
+    if(!json_unpack_ex(j, &error, 0, "{ss*si}", "foo", &s, "baz", &i1))
+        fail("json_unpack failed to catch ! in the middle of an object");
+    check_error("Expected '}' after '*', got 's'", "<format>", 1, 5, 5);
+    json_decref(j);
+
+    /* Error in nested object */
+    j = json_pack("{s{snsn}}", "foo", "bar", "baz");
+    if(!json_unpack_ex(j, &error, 0, "{s{sn!}}", "foo", "bar"))
+        fail("json_unpack nested object with strict validation failed");
+    check_error("1 object item(s) left unpacked", "<validation>", 1, 7, 7);
+    json_decref(j);
+
+    /* Error in nested array */
+    j = json_pack("[[ii]]", 1, 2);
+    if(!json_unpack_ex(j, &error, 0, "[[i!]]", &i1))
+        fail("json_unpack nested array with strict validation failed");
+    check_error("1 array item(s) left unpacked", "<validation>", 1, 5, 5);
+    json_decref(j);
+}

test/suites/api/util.h

@@ -1,20 +1,74 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
  *
  * Jansson is free software; you can redistribute it and/or modify
  * it under the terms of the MIT license. See LICENSE for details.
  */
 
-#ifndef TESTPROGS_UTIL_H
-#define TESTPROGS_UTIL_H
+#ifndef UTIL_H
+#define UTIL_H
 
+#ifdef HAVE_CONFIG_H
+#include <config.h>
+#endif
+
+#include <stdio.h>
 #include <stdlib.h>
+#if HAVE_LOCALE_H
+#include <locale.h>
+#endif
+
+#include <jansson.h>
+
+#define failhdr fprintf(stderr, "%s:%s:%d: ", __FILE__, __FUNCTION__, __LINE__)
 
 #define fail(msg)                                                \
     do {                                                         \
-        fprintf(stderr, "%s:%s:%d: %s\n",                        \
-                __FILE__, __FUNCTION__, __LINE__, msg);          \
+        failhdr;                                                 \
+        fprintf(stderr, "%s\n", msg);                            \
         exit(1);                                                 \
     } while(0)
 
+/* Assumes json_error_t error */
+#define check_error(text_, source_, line_, column_, position_)          \
+    do {                                                                \
+        if(strcmp(error.text, text_) != 0) {                            \
+            failhdr;                                                    \
+            fprintf(stderr, "text: \"%s\" != \"%s\"\n", error.text, text_); \
+            exit(1);                                                    \
+        }                                                               \
+        if(strcmp(error.source, source_) != 0) {                        \
+            failhdr;                                                    \
+                                                                        \
+            fprintf(stderr, "source: \"%s\" != \"%s\"\n", error.source, source_); \
+            exit(1);                                                    \
+        }                                                               \
+        if(error.line != line_) {                                       \
+            failhdr;                                                    \
+            fprintf(stderr, "line: %d != %d\n", error.line, line_);     \
+            exit(1);                                                    \
+        }                                                               \
+        if(error.column != column_) {                                   \
+            failhdr;                                                    \
+            fprintf(stderr, "column: %d != %d\n", error.column, column_); \
+            exit(1);                                                    \
+        }                                                               \
+        if(error.position != position_) {                               \
+            failhdr;                                                    \
+            fprintf(stderr, "position: %d != %d\n", error.position, position_); \
+            exit(1);                                                    \
+        }                                                               \
+    } while(0)
+
+
+static void run_tests();
+
+int main() {
+#ifdef HAVE_SETLOCALE
+    setlocale(LC_ALL, "");
+#endif
+    run_tests();
+    return 0;
+}
+
 #endif

test/suites/encoding-flags/preserve-order/env

@@ -0,0 +1 @@
+export JSON_PRESERVE_ORDER=1

test/suites/encoding-flags/preserve-order/input

@@ -0,0 +1 @@
+{"foo": 1, "bar": 2, "asdf": 3, "deadbeef": 4, "badc0ffee": 5, "qwerty": 6}

test/suites/encoding-flags/preserve-order/output

@@ -0,0 +1 @@
+{"foo": 1, "bar": 2, "asdf": 3, "deadbeef": 4, "badc0ffee": 5, "qwerty": 6}

test/suites/encoding-flags/run

@@ -1,6 +1,6 @@
 #!/bin/sh
 #
-# Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+# Copyright (c) 2009-2011 Petri Lehtinen <petri@digip.org>
 #
 # Jansson is free software; you can redistribute it and/or modify
 # it under the terms of the MIT license. See LICENSE for details.

test/suites/invalid-strip/apostrophe/error

@@ -1,2 +0,0 @@
-1
-invalid token near '''

test/suites/invalid-strip/apostrophe/input

@@ -1 +0,0 @@
-['

test/suites/invalid-strip/ascii-unicode-identifier/error

@@ -1,2 +0,0 @@
-1
-'[' or '{' expected near 'a'

test/suites/invalid-strip/ascii-unicode-identifier/input

@@ -1 +0,0 @@
-aå

test/suites/invalid-strip/brace-comma/error

@@ -1,2 +0,0 @@
-1
-string or '}' expected near ','

test/suites/invalid-strip/brace-comma/input

@@ -1 +0,0 @@
-{,

test/suites/invalid-strip/bracket-comma/error

@@ -1,2 +0,0 @@
-1
-unexpected token near ','

test/suites/invalid-strip/bracket-comma/input

@@ -1 +0,0 @@
-[,

test/suites/invalid-strip/bracket-one-comma/input

@@ -1 +0,0 @@
-[1,

test/suites/invalid-strip/empty/error

@@ -1,2 +0,0 @@
-1
-'[' or '{' expected near end of file

test/suites/invalid-strip/extra-comma-in-array/input

@@ -1 +0,0 @@
-[1,]

test/suites/invalid-strip/extra-command-in-multiline-array/error

@@ -1,2 +0,0 @@
-6
-unexpected token near ']'

test/suites/invalid-strip/extra-command-in-multiline-array/input

@@ -1,6 +0,0 @@
-[1,
-2,
-3,
-4,
-5,
-]

test/suites/invalid-strip/garbage-after-newline/error

@@ -1,2 +0,0 @@
-2
-end of file expected near 'foo'

test/suites/invalid-strip/garbage-after-newline/input

@@ -1,2 +0,0 @@
-[1,2,3]
-foo

test/suites/invalid-strip/garbage-at-the-end/error

@@ -1,2 +0,0 @@
-1
-end of file expected near 'foo'

test/suites/invalid-strip/garbage-at-the-end/input

@@ -1 +0,0 @@
-[1,2,3]foo

test/suites/invalid-strip/integer-starting-with-zero/error

@@ -1,2 +0,0 @@
-1
-invalid token near '0'

test/suites/invalid-strip/integer-starting-with-zero/input

@@ -1 +0,0 @@
-[012]

test/suites/invalid-strip/invalid-escape/error

@@ -1,2 +0,0 @@
-1
-invalid escape near '"\'

test/suites/invalid-strip/invalid-escape/input

@@ -1 +0,0 @@
-["\a <-- invalid escape"]

test/suites/invalid-strip/invalid-identifier/error

@@ -1,2 +0,0 @@
-1
-invalid token near 'troo'

test/suites/invalid-strip/invalid-identifier/input

@@ -1 +0,0 @@
-[troo

test/suites/invalid-strip/invalid-negative-integerr/input

@@ -1 +0,0 @@
-[-123foo]

test/suites/invalid-strip/invalid-negative-real/error

@@ -1,2 +0,0 @@
-1
-']' expected near 'foo'

test/suites/invalid-strip/invalid-negative-real/input

@@ -1 +0,0 @@
-[-123.123foo]

test/suites/invalid-strip/invalid-second-surrogate/error

@@ -1,2 +0,0 @@
-1
-invalid Unicode '\uD888\u3210'

test/suites/invalid-strip/invalid-second-surrogate/input

@@ -1 +0,0 @@
-["\uD888\u3210 (first surrogate and invalid second surrogate)"]

test/suites/invalid-strip/lone-open-brace/input

@@ -1 +0,0 @@
-{

test/suites/invalid-strip/lone-open-bracket/input

@@ -1 +0,0 @@
-[

test/suites/invalid-strip/lone-second-surrogate/error

@@ -1,2 +0,0 @@
-1
-invalid Unicode '\uDFAA'

test/suites/invalid-strip/lone-second-surrogate/input

@@ -1 +0,0 @@
-["\uDFAA (second surrogate on it's own)"]

test/suites/invalid-strip/minus-sign-without-number/error

@@ -1,2 +0,0 @@
-1
-invalid token near '-'

test/suites/invalid-strip/minus-sign-without-number/input

@@ -1 +0,0 @@
-[-foo]

test/suites/invalid-strip/negative-integer-starting-with-zero/error

@@ -1,2 +0,0 @@
-1
-invalid token near '-0'

test/suites/invalid-strip/negative-integer-starting-with-zero/input

@@ -1 +0,0 @@
-[-012]