jansson 1.3

No changes to the actual library code.

.gitignore

@@ -1,3 +1,4 @@
+*~
 *.o
 *.a
 .libs
@@ -22,3 +23,5 @@ missing
 *.la
 stamp-h1
 *.pyc
+*.pc
+/src/jansson.h

CHANGES

@@ -1,11 +1,172 @@
-Version 1.0.4, released 2009-10-11
+Version 1.3
+===========
+
+Released 2010-06-13
+
+* 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
+===========
+
+Released 2010-01-21
+
+* New functions:
+
+  - `json_equal()`: Test whether two JSON values are equal
+  - `json_copy()` and `json_deep_copy()`: Make shallow and deep copies
+    of JSON values
+  - Add a version of all functions taking a string argument that
+    doesn't check for valid UTF-8: `json_string_nocheck()`,
+    `json_string_set_nocheck()`, `json_object_set_nocheck()`,
+    `json_object_set_new_nocheck()`
+
+* New encoding flags:
+
+  - ``JSON_SORT_KEYS``: Sort objects by key
+  - ``JSON_ENSURE_ASCII``: Escape all non-ASCII Unicode characters
+  - ``JSON_COMPACT``: Use a compact representation with all unneeded
+    whitespace stripped
+
+* Bug fixes:
+
+  - Revise and unify whitespace usage in encoder: Add spaces between
+    array and object items, never append newline to output.
+  - Remove const qualifier from the ``json_t`` parameter in
+    `json_string_set()`, `json_integer_set()` and `json_real_set`.
+  - Use ``int32_t`` internally for representing Unicode code points
+    (int is not enough on all platforms)
+
+* Other changes:
+
+  - Convert ``CHANGES`` (this file) to reStructured text and add it to
+    HTML documentation
+  - The test system has been refactored. Python is no longer required
+    to run the tests.
+  - Documentation can now be built by invoking ``make html``
+  - Support for pkg-config
+
+
+Version 1.1.3
+=============
+
+Released 2009-12-18
+
+* Encode reals correctly, so that first encoding and then decoding a
+  real always produces the same value
+* Don't export private symbols in ``libjansson.so``
+
+
+Version 1.1.2
+=============
+
+Released 2009-11-08
+
+* Fix a bug where an error message was not produced if the input file
+  could not be opened in `json_load_file()`
+* Fix an assertion failure in decoder caused by a minus sign without a
+  digit after it
+* Remove an unneeded include of ``stdint.h`` in ``jansson.h``
+
+
+Version 1.1.1
+=============
+
+Released 2009-10-26
+
+* All documentation files were not distributed with v1.1; build
+  documentation in make distcheck to prevent this in the future
+* Fix v1.1 release date in ``CHANGES``
+
+
+Version 1.1
+===========
+
+Released 2009-10-20
+
+* API additions and improvements:
+
+  - Extend array and object APIs
+  - Add functions to modify integer, real and string values
+  - Improve argument validation
+  - Use unsigned int instead of ``uint32_t`` for encoding flags
+
+* Enhance documentation
+
+  - Add getting started guide and tutorial
+  - Fix some typos
+  - General clarifications and cleanup
+
+* Check for integer and real overflows and underflows in decoder
+* Make singleton values thread-safe (``true``, ``false`` and ``null``)
+* Enhance circular reference handling
+* Don't define ``-std=c99`` in ``AM_CFLAGS``
+* Add C++ guards to ``jansson.h``
+* Minor performance and portability improvements
+* Expand test coverage
+
+
+Version 1.0.4
+=============
+
+Released 2009-10-11
 
 * Relax Autoconf version requirement to 2.59
-* Make Jansson compile on platforms where plain char is unsigned
+* Make Jansson compile on platforms where plain ``char`` is unsigned
 * Fix API tests for object
 
 
-Version 1.0.3, released 2009-09-14
+Version 1.0.3
+=============
+
+Released 2009-09-14
 
 * Check for integer and real overflows and underflows in decoder
 * Use the Python json module for tests, or simplejson if the json
@@ -13,16 +174,25 @@ Version 1.0.3, released 2009-09-14
 * Distribute changelog (this file)
 
 
-Version 1.0.2, released 2009-09-08
+Version 1.0.2
+=============
+
+Released 2009-09-08
 
 * Handle EOF correctly in decoder
 
 
-Version 1.0.1, released 2009-09-04
+Version 1.0.1
+=============
 
-* Fixed broken json_is_boolean()
+Released 2009-09-04
+
+* Fixed broken `json_is_boolean()`
 
 
-Version 1.0, released 2009-08-25
+Version 1.0
+===========
+
+Released 2009-08-25
 
 * Initial release

LICENSE

@@ -1,4 +1,4 @@
-Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+Copyright (c) 2009, 2010 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

@@ -1,2 +1,7 @@
 EXTRA_DIST = CHANGES LICENSE README.rst
 SUBDIRS = doc src test
+
+check-local: html
+
+pkgconfigdir = $(libdir)/pkgconfig
+pkgconfig_DATA = jansson.pc

README.rst

@@ -6,14 +6,14 @@ data. Its main features and design principles are:
 
 - Simple and intuitive API and data model
 
-- Good documentation
+- Comprehensive documentation
+
+- No dependencies on other libraries
 
 - Full Unicode support (UTF-8)
 
 - Extensive test suite
 
-- No dependencies on other libraries
-
 Jansson is licensed under the `MIT license`_; see LICENSE in the
 source distribution for details.
 
@@ -36,8 +36,6 @@ To run the test suite, invoke::
 
    $ make check
 
-Python_ is required to run the tests.
-
 
 Documentation
 -------------
@@ -49,15 +47,13 @@ available at http://www.digip.org/jansson/doc/.
 
 To generate HTML documentation yourself, invoke::
 
-   cd doc/
-   sphinx-build . .build/html
+   make html
 
-... and point your browser to ``.build/html/index.html``. Sphinx_ is
+and point your browser to ``doc/_build/html/index.html``. Sphinx_ is
 required to generate the documentation.
 
 
 .. _Jansson: http://www.digip.org/jansson/
 .. _`MIT license`: http://www.opensource.org/licenses/mit-license.php
-.. _Python: http://www.python.org/
 .. _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], [1.0.4], [petri@digip.org])
+AC_PREREQ([2.60])
+AC_INIT([jansson], [1.3], [petri@digip.org])
 
 AM_INIT_AUTOMAKE([1.10 foreign])
 
@@ -8,6 +8,7 @@ AC_CONFIG_HEADERS([config.h])
 
 # Checks for programs.
 AC_PROG_CC
+AC_PROG_CXX
 AC_PROG_LIBTOOL
 
 # Checks for libraries.
@@ -15,15 +16,27 @@ AC_PROG_LIBTOOL
 # Checks for header files.
 
 # Checks for typedefs, structures, and compiler characteristics.
+AC_TYPE_INT32_T
+
+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_CONFIG_FILES([
+        jansson.pc
         Makefile
         doc/Makefile
         src/Makefile
+        src/jansson.h
         test/Makefile
-        test/testdata/Makefile
-        test/testprogs/Makefile
+        test/bin/Makefile
+        test/suites/Makefile
+        test/suites/api/Makefile
 ])
 AC_OUTPUT

doc/.gitignore

@@ -1 +1 @@
-.build/
+_build/

doc/Makefile.am

@@ -1,4 +1,20 @@
-EXTRA_DIST = conf.py apiref.rst index.rst ext/refcounting.py
+EXTRA_DIST = conf.py apiref.rst changes.rst conformance.rst		\
+	gettingstarted.rst github_commits.c index.rst tutorial.rst	\
+	ext/refcounting.py
+
+SPHINXBUILD = sphinx-build
+SPHINXOPTS = -d _build/doctrees -W
+
+html-local:
+	$(SPHINXBUILD) -b html $(SPHINXOPTS) $(srcdir) _build/html
+
+install-html-local: html
+	mkdir -p $(DESTDIR)$(htmldir)
+	cp -r _build/html $(DESTDIR)$(htmldir)
+
+uninstall-local:
+	rm -rf $(DESTDIR)$(htmldir)
 
 clean-local:
-	rm -rf .build
+	rm -rf _build
+	rm -f ext/refcounting.pyc

doc/README

@@ -1,5 +1,5 @@
 To build the documentation, invoke
 
-    sphinx-build . .build/html
+    make html
 
-in this directory. Then point your browser to .build/html/index.html.
+Then point your browser to _build/html/index.html.

doc/apiref.rst

@@ -1,3 +1,5 @@
+.. _apiref:
+
 *************
 API Reference
 *************
@@ -41,6 +43,12 @@ Objects of :ctype:`json_t` are always used through a pointer. There
 are APIs for querying the type, manipulating the reference count, and
 for constructing and manipulating values of different types.
 
+Unless noted otherwise, all API functions return an error value if an
+error occurs. Depending on the function's signature, the error value
+is either *NULL* or -1. Invalid arguments or invalid input are
+apparent sources for errors. Memory allocation and I/O operations may
+also cause errors.
+
 
 Type
 ----
@@ -80,8 +88,8 @@ functions:
 .. cfunction:: int json_typeof(const json_t *json)
 
    Return the type of the JSON value (a :ctype:`json_type` cast to
-   :ctype:`int`). This function is actually implemented as a macro for
-   speed.
+   :ctype:`int`). *json* MUST NOT be *NULL*. This function is actually
+   implemented as a macro for speed.
 
 .. cfunction:: json_is_object(const json_t *json)
                json_is_array(const json_t *json)
@@ -93,19 +101,22 @@ functions:
                json_is_null(const json_t *json)
 
    These functions (actually macros) return true (non-zero) for values
-   of the given type, and false (zero) for values of other types.
+   of the given type, and false (zero) for values of other types and
+   for *NULL*.
 
 .. cfunction:: json_is_number(const json_t *json)
 
    Returns true for values of types :const:`JSON_INTEGER` and
-   :const:`JSON_REAL`, and false for other types.
+   :const:`JSON_REAL`, and false for other types and for *NULL*.
 
 .. cfunction:: json_is_boolean(const json_t *json)
 
    Returns true for types :const:`JSON_TRUE` and :const:`JSON_FALSE`,
-   and false for values of other types.
+   and false for values of other types and for *NULL*.
 
 
+.. _apiref-reference-count:
+
 Reference Count
 ---------------
 
@@ -121,7 +132,8 @@ The following functions are used to manipulate the reference count.
 
 .. cfunction:: json_t *json_incref(json_t *json)
 
-   Increment the reference count of *json*.
+   Increment the reference count of *json* if it's not non-*NULL*.
+   Returns *json*.
 
 .. cfunction:: void json_decref(json_t *json)
 
@@ -142,38 +154,91 @@ Normally, all functions accepting a JSON value as an argument will
 manage the reference, i.e. increase and decrease the reference count
 as needed. However, some functions **steal** the reference, i.e. they
 have the same result as if the user called :cfunc:`json_decref()` on
-the argument right after calling the function. These are usually
-convenience functions for adding new references to containers and not
-to worry about the reference count.
+the argument right after calling the function. These functions are
+suffixed with ``_new`` or have ``_new_`` somewhere in their name.
+
+For example, the following code creates a new JSON array and appends
+an integer to it::
+
+  json_t *array, *integer;
+
+  array = json_array();
+  integer = json_integer(42);
+
+  json_array_append(array, integer);
+  json_decref(integer);
+
+Note how the caller has to release the reference to the integer value
+by calling :cfunc:`json_decref()`. By using a reference stealing
+function :cfunc:`json_array_append_new()` instead of
+:cfunc:`json_array_append()`, the code becomes much simpler::
+
+  json_t *array = json_array();
+  json_array_append_new(array, json_integer(42));
+
+In this case, the user doesn't have to explicitly release the
+reference to the integer value, as :cfunc:`json_array_append_new()`
+steals the reference when appending the value to the array.
 
 In the following sections it is clearly documented whether a function
 will return a new or borrowed reference or steal a reference to its
 argument.
 
 
+Circular References
+-------------------
+
+A circular reference is created when an object or an array is,
+directly or indirectly, inserted inside itself. The direct case is
+simple::
+
+  json_t *obj = json_object();
+  json_object_set(obj, "foo", obj);
+
+Jansson will refuse to do this, and :cfunc:`json_object_set()` (and
+all the other such functions for objects and arrays) will return with
+an error status. The indirect case is the dangerous one::
+
+  json_t *arr1 = json_array(), *arr2 = json_array();
+  json_array_append(arr1, arr2);
+  json_array_append(arr2, arr1);
+
+In this example, the array ``arr2`` is contained in the array
+``arr1``, and vice versa. Jansson cannot check for this kind of
+indirect circular references without a performance hit, so it's up to
+the user to avoid them.
+
+If a circular reference is created, the memory consumed by the values
+cannot be freed by :cfunc:`json_decref()`. The reference counts never
+drops to zero because the values are keeping the circular reference to
+themselves. Moreover, trying to encode the values with any of the
+encoding functions will fail. The encoder detects circular references
+and returns an error status.
+
+
 True, False and Null
 ====================
 
+These values are implemented as singletons, so each of these functions
+returns the same value each time.
+
 .. cfunction:: json_t *json_true(void)
 
    .. refcounting:: new
 
-   Returns a value of the type :const:`JSON_TRUE`, or *NULL* on
-   error.
+   Returns the JSON true value.
 
 .. cfunction:: json_t *json_false(void)
 
    .. refcounting:: new
 
-   Returns a value of the type :const:`JSON_FALSE`, or *NULL* on
-   error.
+   Returns the JSON false value.
 
 .. cfunction:: json_t *json_null(void)
 
    .. refcounting:: new
 
-   Returns a value of the type :const:`JSON_NULL`, or *NULL* on
-   error.
+   Returns the JSON null value.
 
 
 String
@@ -183,13 +248,40 @@ String
 
    .. refcounting:: new
 
-   Returns a new value of the type :const:`JSON_STRING`, or *NULL* on
-   error. *value* must be a valid UTF-8 encoded Unicode string.
+   Returns a new JSON string, or *NULL* on error. *value* must be a
+   valid UTF-8 encoded Unicode string.
 
-.. cfunction:: const char *json_string_value(const json_t *json)
+.. cfunction:: json_t *json_string_nocheck(const char *value)
 
-   Returns the associated value of a :const:`JSON_STRING` value as a
-   null terminated UTF-8 encoded string.
+   .. refcounting:: new
+
+   Like :cfunc:`json_string`, but doesn't check that *value* is valid
+   UTF-8. Use this function only if you are certain that this really
+   is the case (e.g. you have already checked it by other means).
+
+   .. versionadded:: 1.2
+
+.. cfunction:: const char *json_string_value(const json_t *string)
+
+   Returns the associated value of *string* as a null terminated UTF-8
+   encoded string, or *NULL* if *string* is not a JSON string.
+
+.. cfunction:: int json_string_set(const json_t *string, const char *value)
+
+   Sets the associated value of *string* to *value*. *value* must be a
+   valid UTF-8 encoded Unicode string. Returns 0 on success and -1 on
+   error.
+
+   .. versionadded:: 1.1
+
+.. cfunction:: int json_string_set_nocheck(const json_t *string, const char *value)
+
+   Like :cfunc:`json_string_set`, but doesn't check that *value* is
+   valid UTF-8. Use this function only if you are certain that this
+   really is the case (e.g. you have already checked it by other
+   means).
+
+   .. versionadded:: 1.2
 
 
 Number
@@ -199,33 +291,46 @@ Number
 
    .. refcounting:: new
 
-   Returns a new value of the type :const:`JSON_INTEGER`, or *NULL* on
-   error.
+   Returns a new JSON integer, or *NULL* on error.
 
-.. cfunction:: int json_integer_value(const json_t *json)
+.. cfunction:: int json_integer_value(const json_t *integer)
 
-   Returns the associated integer value of values of the type
-   :const:`JSON_INTEGER`, or 0 for values of other types.
+   Returns the associated value of *integer*, or 0 if *json* is not a
+   JSON integer.
+
+.. cfunction:: int json_integer_set(const json_t *integer, int value)
+
+   Sets the associated value of *integer* to *value*. Returns 0 on
+   success and -1 if *integer* is not a JSON integer.
+
+   .. versionadded:: 1.1
 
 .. cfunction:: json_t *json_real(double value)
 
    .. refcounting:: new
 
-   Returns a new value of the type :const:`JSON_REAL`, or *NULL* on
-   error.
+   Returns a new JSON real, or *NULL* on error.
 
-.. cfunction:: double json_real_value(const json_t *json)
+.. cfunction:: double json_real_value(const json_t *real)
 
-   Returns the associated real value of values of the type
-   :const:`JSON_INTEGER`, or 0 for values of other types.
+   Returns the associated value of *real*, or 0.0 if *real* is not a
+   JSON real.
+
+.. cfunction:: int json_real_set(const json_t *real, double value)
+
+   Sets the associated value of *real* to *value*. Returns 0 on
+   success and -1 if *real* is not a JSON real.
+
+   .. versionadded:: 1.1
 
 In addition to the functions above, there's a common query function
 for integers and reals:
 
 .. cfunction:: double json_number_value(const json_t *json)
 
-   Returns the value of either ``JSON_INTEGER`` or ``JSON_REAL``, cast
-   to double regardless of the actual type.
+   Returns the associated value of the JSON integer or JSON real
+   *json*, cast to double regardless of the actual type. If *json* is
+   neither JSON real nor JSON integer, 0.0 is returned.
 
 
 Array
@@ -237,33 +342,90 @@ A JSON array is an ordered collection of other JSON values.
 
    .. refcounting:: new
 
-   Returns a new value of the type :const:`JSON_ARRAY`, or *NULL* on
-   error. Initially, the array is empty.
+   Returns a new JSON array, or *NULL* on error. Initially, the array
+   is empty.
 
 .. cfunction:: unsigned int json_array_size(const json_t *array)
 
-   Returns the number of elements in *array*.
+   Returns the number of elements in *array*, or 0 if *array* is NULL
+   or not a JSON array.
 
 .. cfunction:: json_t *json_array_get(const json_t *array, unsigned int index)
 
    .. refcounting:: borrow
 
-   Returns the element in *array* at position *index*, or *NULL* if
-   *index* is out of range. The valid range for *index* is from 0 to
-   the return value of :cfunc:`json_array_size()` minus 1.
+   Returns the element in *array* at position *index*. The valid range
+   for *index* is from 0 to the return value of
+   :cfunc:`json_array_size()` minus 1. If *array* is not a JSON array,
+   if *array* is *NULL*, or if *index* is out of range, *NULL* is
+   returned.
 
 .. cfunction:: int json_array_set(json_t *array, unsigned int index, json_t *value)
 
    Replaces the element in *array* at position *index* with *value*.
-   Returns 0 on success, or -1 if *index* is out of range. The valid
-   range for *index* is from 0 to the return value of
-   :cfunc:`json_array_size()` minus 1.
+   The valid range for *index* is from 0 to the return value of
+   :cfunc:`json_array_size()` minus 1. Returns 0 on success and -1 on
+   error.
+
+.. cfunction:: int json_array_set_new(json_t *array, unsigned int index, json_t *value)
+
+   Like :cfunc:`json_array_set()` but steals the reference to *value*.
+   This is useful when *value* is newly created and not used after
+   the call.
+
+   .. versionadded:: 1.1
 
 .. cfunction:: int json_array_append(json_t *array, json_t *value)
 
    Appends *value* to the end of *array*, growing the size of *array*
    by 1. Returns 0 on success and -1 on error.
 
+.. cfunction:: int json_array_append_new(json_t *array, json_t *value)
+
+   Like :cfunc:`json_array_append()` but steals the reference to
+   *value*. This is useful when *value* is newly created and not used
+   after the call.
+
+   .. versionadded:: 1.1
+
+.. cfunction:: int json_array_insert(json_t *array, unsigned int index, json_t *value)
+
+   Inserts *value* to *array* at position *index*, shifting the
+   elements at *index* and after it one position towards the end of
+   the array. Returns 0 on success and -1 on error.
+
+   .. versionadded:: 1.1
+
+.. cfunction:: int json_array_insert_new(json_t *array, unsigned int index, json_t *value)
+
+   Like :cfunc:`json_array_insert()` but steals the reference to
+   *value*. This is useful when *value* is newly created and not used
+   after the call.
+
+   .. versionadded:: 1.1
+
+.. cfunction:: int json_array_remove(json_t *array, unsigned int index)
+
+   Removes the element in *array* at position *index*, shifting the
+   elements after *index* one position towards the start of the array.
+   Returns 0 on success and -1 on error.
+
+   .. versionadded:: 1.1
+
+.. cfunction:: int json_array_clear(json_t *array)
+
+   Removes all elements from *array*. Returns 0 on sucess and -1 on
+   error.
+
+   .. versionadded:: 1.1
+
+.. cfunction:: int json_array_extend(json_t *array, json_t *other_array)
+
+   Appends all elements in *other_array* to the end of *array*.
+   Returns 0 on success and -1 on error.
+
+   .. versionadded:: 1.1
+
 
 Object
 ======
@@ -275,8 +437,15 @@ Unicode string and the value is any JSON value.
 
    .. refcounting:: new
 
-   Returns a new value of the type :const:`JSON_OBJECT`, or *NULL* on
-   error. Initially, the object is empty.
+   Returns a new JSON object, or *NULL* on error. Initially, the
+   object is empty.
+
+.. cfunction:: unsigned int json_object_size(const json_t *object)
+
+   Returns the number of elements in *object*, or 0 if *object* is not
+   a JSON object.
+
+   .. versionadded:: 1.1
 
 .. cfunction:: json_t *json_object_get(const json_t *object, const char *key)
 
@@ -288,9 +457,35 @@ Unicode string and the value is any JSON value.
 .. cfunction:: int json_object_set(json_t *object, const char *key, json_t *value)
 
    Set the value of *key* to *value* in *object*. *key* must be a
-   valid terminated UTF-8 encoded Unicode string. If there already is
-   a value for *key*, it is replaced by the new value. Returns 0 on
-   success and -1 on error.
+   valid null terminated UTF-8 encoded Unicode string. If there
+   already is a value for *key*, it is replaced by the new value.
+   Returns 0 on success and -1 on error.
+
+.. cfunction:: int json_object_set_nocheck(json_t *object, const char *key, json_t *value)
+
+   Like :cfunc:`json_object_set`, but doesn't check that *key* is
+   valid UTF-8. Use this function only if you are certain that this
+   really is the case (e.g. you have already checked it by other
+   means).
+
+   .. versionadded:: 1.2
+
+.. cfunction:: int json_object_set_new(json_t *object, const char *key, json_t *value)
+
+   Like :cfunc:`json_object_set()` but steals the reference to
+   *value*. This is useful when *value* is newly created and not used
+   after the call.
+
+   .. versionadded:: 1.1
+
+.. cfunction:: int json_object_set_new_nocheck(json_t *object, const char *key, json_t *value)
+
+   Like :cfunc:`json_object_set_new`, but doesn't check that *key* is
+   valid UTF-8. Use this function only if you are certain that this
+   really is the case (e.g. you have already checked it by other
+   means).
+
+   .. versionadded:: 1.2
 
 .. cfunction:: int json_object_del(json_t *object, const char *key)
 
@@ -298,6 +493,21 @@ Unicode string and the value is any JSON value.
    -1 if *key* was not found.
 
 
+.. cfunction:: int json_object_clear(json_t *object)
+
+   Remove all elements from *object*. Returns 0 on success and -1 if
+   *object* is not a JSON object.
+
+   .. versionadded:: 1.1
+
+.. cfunction:: int json_object_update(json_t *object, json_t *other)
+
+   Update *object* with the key-value pairs from *other*, overwriting
+   existing keys. Returns 0 on success or -1 on error.
+
+   .. versionadded:: 1.1
+
+
 The following functions implement an iteration protocol for objects:
 
 .. cfunction:: void *json_object_iter(json_t *object)
@@ -305,6 +515,16 @@ The following functions implement an iteration protocol for objects:
    Returns an opaque iterator which can be used to iterate over all
    key-value pairs in *object*, or *NULL* if *object* is empty.
 
+.. cfunction:: void *json_object_iter_at(json_t *object, const char *key)
+
+   Like :cfunc:`json_object_iter()`, but returns an iterator to the
+   key-value pair in *object* whose key is equal to *key*, or NULL if
+   *key* is not found in *object*. Iterating forward to the end of
+   *object* only yields all key-value pairs of the object if *key*
+   happens to be the first key in the underlying hash table.
+
+   .. versionadded:: 1.3
+
 .. cfunction:: void *json_object_iter_next(json_t *object, void *iter)
 
    Returns an iterator pointing to the next key-value pair in *object*
@@ -321,6 +541,35 @@ The following functions implement an iteration protocol for objects:
 
    Extract the associated value from *iter*.
 
+.. cfunction:: int json_object_iter_set(json_t *object, void *iter, json_t *value)
+
+   Set the value of the key-value pair in *object*, that is pointed to
+   by *iter*, to *value*.
+
+   .. versionadded:: 1.3
+
+.. cfunction:: int json_object_iter_set_new(json_t *object, void *iter, json_t *value)
+
+   Like :cfunc:`json_object_iter_set()`, but steals the reference to
+   *value*. This is useful when *value* is newly created and not used
+   after the call.
+
+   .. versionadded:: 1.3
+
+The iteration protocol can be used for example as follows::
+
+   /* obj is a JSON object */
+   const char *key;
+   json_t *value;
+   void *iter = json_object_iter(obj);
+   while(iter)
+   {
+       key = json_object_iter_key(iter);
+       value = json_object_iter_value(iter);
+       /* use key and value ... */
+       iter = json_object_iter_next(obj, iter);
+   }
+
 
 Encoding
 ========
@@ -329,32 +578,71 @@ This sections describes the functions that can be used to encode
 values to JSON. Only objects and arrays can be encoded, since they are
 the only valid "root" values of a JSON text.
 
+By default, the output has no newlines, and spaces are used between
+array and object elements for a readable output. This behavior can be
+altered by using the ``JSON_INDENT`` and ``JSON_COMPACT`` flags
+described below. A newline is never appended to the end of the encoded
+JSON data.
+
 Each function takes a *flags* parameter that controls some aspects of
 how the data is encoded. Its default value is 0. The following macros
 can be ORed together to obtain *flags*.
 
 ``JSON_INDENT(n)``
-   Pretty-print the result, indenting arrays and objects by *n*
-   spaces. The valid range for *n* is between 0 and 255, other values
-   result in an undefined output. If ``JSON_INDENT`` is not used or
-   *n* is 0, no pretty-printing is done and the result is a compact
-   representation.
+   Pretty-print the result, using newlines between array and object
+   items, and indenting with *n* spaces. The valid range for *n* is
+   between 0 and 255, other values result in an undefined output. If
+   ``JSON_INDENT`` is not used or *n* is 0, no newlines are inserted
+   between array and object items.
+
+``JSON_COMPACT``
+   This flag enables a compact representation, i.e. sets the separator
+   between array and object items to ``","`` and between object keys
+   and values to ``":"``. Without this flag, the corresponding
+   separators are ``", "`` and ``": "`` for more readable output.
+
+   .. versionadded:: 1.2
+
+``JSON_ENSURE_ASCII``
+   If this flag is used, the output is guaranteed to consist only of
+   ASCII characters. This is achived by escaping all Unicode
+   characters outside the ASCII range.
+
+   .. versionadded:: 1.2
+
+``JSON_SORT_KEYS``
+   If this flag is used, all the objects in output are sorted by key.
+   This is useful e.g. if two JSON texts are diffed or visually
+   compared.
+
+   .. versionadded:: 1.2
+
+``JSON_PRESERVE_ORDER``
+   If this flag is used, object keys in the output are sorted into the
+   same order in which they were first inserted to the object. For
+   example, decoding a JSON text and then encoding with this flag
+   preserves the order of object keys.
+
+   .. versionadded:: 1.3
 
 The following functions perform the actual JSON encoding. The result
 is in UTF-8.
 
-.. cfunction:: char *json_dumps(const json_t *root, uint32_t flags)
+.. cfunction:: char *json_dumps(const json_t *root, unsigned long flags)
 
    Returns the JSON representation of *root* as a string, or *NULL* on
    error. *flags* is described above. The return value must be freed
    by the caller using :cfunc:`free()`.
 
-.. cfunction:: int json_dumpf(const json_t *root, FILE *output, uint32_t flags)
+.. cfunction:: int json_dumpf(const json_t *root, FILE *output, unsigned long flags)
 
    Write the JSON representation of *root* to the stream *output*.
    *flags* is described above. Returns 0 on success and -1 on error.
+   If an error occurs, something may have already been written to
+   *output*. In this case, the output is undefined and most likely not
+   valid JSON.
 
-.. cfunction:: int json_dump_file(const json_t *json, const char *path, uint32_t flags)
+.. cfunction:: int json_dump_file(const json_t *json, const char *path, unsigned long flags)
 
    Write the JSON representation of *root* to the file *path*. If
    *path* already exists, it is overwritten. *flags* is described
@@ -440,3 +728,75 @@ The following functions perform the actual JSON decoding.
    object it contains, or *NULL* on error, in which case *error* is
    filled with information about the error. See above for discussion
    on the *error* parameter.
+
+
+Equality
+========
+
+Testing for equality of two JSON values cannot, in general, be
+achieved using the ``==`` operator. Equality in the terms of the
+``==`` operator states that the two :ctype:`json_t` pointers point to
+exactly the same JSON value. However, two JSON values can be equal not
+only if they are exactly the same value, but also if they have equal
+"contents":
+
+* Two integer or real values are equal if their contained numeric
+  values are equal. An integer value is never equal to a real value,
+  though.
+
+* Two strings are equal if their contained UTF-8 strings are equal.
+
+* Two arrays are equal if they have the same number of elements and
+  each element in the first array is equal to the corresponding
+  element in the second array.
+
+* Two objects are equal if they have exactly the same keys and the
+  value for each key in the first object is equal to the value of the
+  corresponding key in the second object.
+
+* Two true, false or null values have no "contents", so they are equal
+  if their types are equal. (Because these values are singletons,
+  their equality can actually be tested with ``==``.)
+
+The following function can be used to test whether two JSON values are
+equal.
+
+.. cfunction:: int json_equal(json_t *value1, json_t *value2)
+
+   Returns 1 if *value1* and *value2* are equal, as defined above.
+   Returns 0 if they are inequal or one or both of the pointers are
+   *NULL*.
+
+   .. versionadded:: 1.2
+
+
+Copying
+=======
+
+Because of reference counting, passing JSON values around doesn't
+require copying them. But sometimes a fresh copy of a JSON value is
+needed. For example, if you need to modify an array, but still want to
+use the original afterwards, you should take a copy of it first.
+
+Jansson supports two kinds of copying: shallow and deep. There is a
+difference between these methods only for arrays and objects. Shallow
+copying only copies the first level value (array or object) and uses
+the same child values in the copied value. Deep copying makes a fresh
+copy of the child values, too. Moreover, all the child values are deep
+copied in a recursive fashion.
+
+.. cfunction:: json_t *json_copy(json_t *value)
+
+   .. refcounting:: new
+
+   Returns a shallow copy of *value*, or *NULL* on error.
+
+   .. versionadded:: 1.2
+
+.. cfunction:: json_t *json_deep_copy(json_t *value)
+
+   .. refcounting:: new
+
+   Returns a deep copy of *value*, or *NULL* on error.
+
+   .. versionadded:: 1.2

doc/changes.rst

@@ -0,0 +1,5 @@
+******************
+Changes in Jansson
+******************
+
+.. include:: ../CHANGES

doc/conf.py

@@ -43,16 +43,16 @@ master_doc = 'index'
 
 # General information about the project.
 project = u'Jansson'
-copyright = u'2009, Petri Lehtinen'
+copyright = u'2009, 2010 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 = '1.0'
+version = '1.3'
 # The full version, including alpha/beta/rc tags.
-release = '1.0.4'
+release = '1.3'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.
@@ -69,10 +69,10 @@ release = '1.0.4'
 
 # List of directories, relative to source directory, that shouldn't be searched
 # for source files.
-exclude_trees = ['.build']
+exclude_trees = ['_build']
 
 # The reST default role (used for this markup: `text`) to use for all documents.
-#default_role = None
+default_role = 'cfunc'
 
 # If true, '()' will be appended to :func: etc. cross-reference text.
 #add_function_parentheses = True

doc/conformance.rst

@@ -0,0 +1,102 @@
+***************
+RFC Conformance
+***************
+
+JSON is specified in :rfc:`4627`, *"The application/json Media Type
+for JavaScript Object Notation (JSON)"*. This chapter discusses
+Jansson's conformance to this specification.
+
+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.
+
+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 ``int`` type.
+
+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`` type 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 ``int`` type 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
+``int`` and ``double``. This excludes things such as unsigned types,
+``long``, ``long long``, ``long double``, etc. Obviously, shorter
+types like ``short`` 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, 2010 Petri Lehtinen <petri@digip.org>
     :license: MIT, see LICENSE for details.
 """
 

doc/gettingstarted.rst

@@ -0,0 +1,108 @@
+***************
+Getting Started
+***************
+
+.. highlight:: c
+
+Compiling and Installing Jansson
+================================
+
+The Jansson source is available at
+http://www.digip.org/jansson/releases/.
+
+Unpack the source tarball and change to the source directory:
+
+.. parsed-literal::
+
+    bunzip2 -c jansson-|release|.tar.bz2 | tar xf -
+    cd jansson-|release|
+
+The source uses GNU Autotools (autoconf_, automake_, libtool_), so
+compiling and installing is extremely simple::
+
+    ./configure
+    make
+    make check
+    make install
+
+To change the destination directory (``/usr/local`` by default), use
+the ``--prefix=DIR`` argument to ``./configure``. See ``./configure
+--help`` for the list of all possible installation options. (There are
+no options to customize the resulting Jansson binary.)
+
+The command ``make check`` runs the test suite distributed with
+Jansson. This step is not strictly necessary, but it may find possible
+problems that Jansson has on your platform. If any problems are found,
+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
+bootstrapped. There are many ways to do this, but the easiest one is
+to use ``autoreconf``::
+
+    autoreconf -vi
+
+This command creates the ``./configure`` script, which can then be
+used as described above.
+
+.. _autoconf: http://www.gnu.org/software/autoconf/
+.. _automake: http://www.gnu.org/software/automake/
+.. _libtool: http://www.gnu.org/software/libtool/
+
+
+Installing Prebuilt Binary Packages
+-----------------------------------
+
+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::
+
+  sudo apt-get install libjansson-dev
+
+.. _this PPA: http://launchpad.net/~petri/+archive/ppa
+.. _Launchpad: http://launchpad.net/
+
+
+Building the Documentation
+--------------------------
+
+(This subsection describes how to build the HTML documentation you are
+currently reading, so it can be safely skipped.)
+
+Documentation is in the ``doc/`` subdirectory. It's written in
+reStructuredText_ with Sphinx_ annotations. To generate the HTML
+documentation, invoke::
+
+   make html
+
+and point your browser to ``doc/_build/html/index.html``. Sphinx_ is
+required to generate the documentation.
+
+.. _reStructuredText: http://docutils.sourceforge.net/rst.html
+.. _Sphinx: http://sphinx.pocoo.org/
+
+
+Compiling Programs That Use Jansson
+===================================
+
+Jansson involves one C header file, :file:`jansson.h`, so it's enough
+to put the line
+
+::
+
+    #include <jansson.h>
+
+in the beginning of every source file that uses Jansson.
+
+There's also just one library to link with, ``libjansson``. Compile and
+link the program as follows::
+
+    cc -o prog prog.c -ljansson
+
+Starting from version 1.2, there's also support for pkg-config_::
+
+    cc -o prog prog.c `pkg-config --cflags --libs jansson`
+
+.. _pkg-config: http://pkg-config.freedesktop.org/

doc/github_commits.c

@@ -0,0 +1,171 @@
+/*
+ * Copyright (c) 2009, 2010 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 <stdlib.h>
+#include <string.h>
+
+#include <jansson.h>
+#include <curl/curl.h>
+
+#define BUFFER_SIZE  (256 * 1024)  /* 256 KB */
+
+#define URL_FORMAT   "http://github.com/api/v2/json/commits/list/%s/%s/master"
+#define URL_SIZE     256
+
+/* Return the offset of the first newline in text or the length of
+   text if there's no newline */
+static int newline_offset(const char *text)
+{
+    const char *newline = strchr(text, '\n');
+    if(!newline)
+        return strlen(text);
+    else
+        return (int)(newline - text);
+}
+
+struct write_result
+{
+    char *data;
+    int pos;
+};
+
+static size_t write_response(void *ptr, size_t size, size_t nmemb, void *stream)
+{
+    struct write_result *result = (struct write_result *)stream;
+
+    if(result->pos + size * nmemb >= BUFFER_SIZE - 1)
+    {
+        fprintf(stderr, "error: too small buffer\n");
+        return 0;
+    }
+
+    memcpy(result->data + result->pos, ptr, size * nmemb);
+    result->pos += size * nmemb;
+
+    return size * nmemb;
+}
+
+static char *request(const char *url)
+{
+    CURL *curl;
+    CURLcode status;
+    char *data;
+    long code;
+
+    curl = curl_easy_init();
+    data = malloc(BUFFER_SIZE);
+    if(!curl || !data)
+        return NULL;
+
+    struct write_result write_result = {
+        .data = data,
+        .pos = 0
+    };
+
+    curl_easy_setopt(curl, CURLOPT_URL, url);
+    curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, write_response);
+    curl_easy_setopt(curl, CURLOPT_WRITEDATA, &write_result);
+
+    status = curl_easy_perform(curl);
+    if(status != 0)
+    {
+        fprintf(stderr, "error: unable to request data from %s:\n", url);
+        fprintf(stderr, "%s\n", curl_easy_strerror(status));
+        return NULL;
+    }
+
+    curl_easy_getinfo(curl, CURLINFO_RESPONSE_CODE, &code);
+    if(code != 200)
+    {
+        fprintf(stderr, "error: server responded with code %ld\n", code);
+        return NULL;
+    }
+
+    curl_easy_cleanup(curl);
+    curl_global_cleanup();
+
+    /* zero-terminate the result */
+    data[write_result.pos] = '\0';
+
+    return data;
+}
+
+int main(int argc, char *argv[])
+{
+    unsigned int i;
+    char *text;
+    char url[URL_SIZE];
+
+    json_t *root;
+    json_error_t error;
+    json_t *commits;
+
+    if(argc != 3)
+    {
+        fprintf(stderr, "usage: %s USER REPOSITORY\n\n", argv[0]);
+        fprintf(stderr, "List commits at USER's REPOSITORY.\n\n");
+        return 2;
+    }
+
+    snprintf(url, URL_SIZE, URL_FORMAT, argv[1], argv[2]);
+
+    text = request(url);
+    if(!text)
+        return 1;
+
+    root = json_loads(text, &error);
+    free(text);
+
+    if(!root)
+    {
+        fprintf(stderr, "error: on line %d: %s\n", error.line, error.text);
+        return 1;
+    }
+
+    commits = json_object_get(root, "commits");
+    if(!json_is_array(commits))
+    {
+        fprintf(stderr, "error: commits is not an array\n");
+        return 1;
+    }
+
+    for(i = 0; i < json_array_size(commits); i++)
+    {
+        json_t *commit, *id, *message;
+        const char *message_text;
+
+        commit = json_array_get(commits, i);
+        if(!json_is_object(commit))
+        {
+            fprintf(stderr, "error: commit %d is not an object\n", i + 1);
+            return 1;
+        }
+
+        id = json_object_get(commit, "id");
+        if(!json_is_string(id))
+        {
+            fprintf(stderr, "error: commit %d: id is not a string\n", i + 1);
+            return 1;
+        }
+
+        message = json_object_get(commit, "message");
+        if(!json_is_string(message))
+        {
+            fprintf(stderr, "error: commit %d: message is not a string\n", i + 1);
+            return 1;
+        }
+
+        message_text = json_string_value(message);
+        printf("%.8s %.*s\n",
+               json_string_value(id),
+               newline_offset(message_text),
+               message_text);
+    }
+
+    json_decref(root);
+    return 0;
+}

doc/index.rst

@@ -1,16 +1,42 @@
-Overview
-========
+Jansson Documentation
+=====================
 
 This is the documentation for Jansson_ |release|, last updated |today|.
 
+Introduction
+------------
+
+Jansson_ is a C library for encoding, decoding and manipulating JSON
+data. Its main features and design principles are:
+
+- Simple and intuitive API and data model
+
+- Comprehensive documentation
+
+- No dependencies on other libraries
+
+- Full Unicode support (UTF-8)
+
+- Extensive test suite
+
+Jansson is licensed under the `MIT license`_; see LICENSE in the
+source distribution for details.
+
+
+.. _`MIT license`: http://www.opensource.org/licenses/mit-license.php
 .. _Jansson: http://www.digip.org/jansson/
 
-**Contents:**
+Contents
+--------
 
 .. toctree::
    :maxdepth: 2
 
+   gettingstarted
+   tutorial
+   conformance
    apiref
+   changes
 
 
 Indices and Tables

doc/tutorial.rst

@@ -0,0 +1,275 @@
+.. _tutorial:
+
+********
+Tutorial
+********
+
+.. highlight:: c
+
+In this tutorial, we create a program that fetches the latest commits
+of a repository in GitHub_ over the web. One of the response formats
+supported by `GitHub API`_ is JSON, so the result can be parsed using
+Jansson.
+
+To stick to the the scope of this tutorial, we will only cover the the
+parts of the program related to handling JSON data. For the best user
+experience, the full source code is available:
+:download:`github_commits.c`. To compile it (on Unix-like systems with
+gcc), use the following command::
+
+    gcc -o github_commits github_commits.c -ljansson -lcurl
+
+libcurl_ is used to communicate over the web, so it is required to
+compile the program.
+
+The command line syntax is::
+
+    github_commits USER REPOSITORY
+
+``USER`` is a GitHub user ID and ``REPOSITORY`` is the repository
+name. Please note that the GitHub API is rate limited, so if you run
+the program too many times within a short period of time, the sever
+starts to respond with an error.
+
+.. _GitHub: http://github.com/
+.. _GitHub API: http://develop.github.com/
+.. _libcurl: http://curl.haxx.se/
+
+
+.. _tutorial-github-commits-api:
+
+The GitHub Commits API
+======================
+
+The `GitHub commits API`_ is used by sending HTTP requests to URLs
+starting with ``http://github.com/api/v2/json/commits/``. Our program
+only lists the latest commits, so the rest of the URL is
+``list/USER/REPOSITORY/BRANCH``, where ``USER``, ``REPOSITORY`` and
+``BRANCH`` are the GitHub user ID, the name of the repository, and the
+name of the branch whose commits are to be listed, respectively.
+
+GitHub responds with a JSON object of the following form:
+
+.. code-block:: none
+
+    {
+        "commits": [
+            {
+                "id": "<the commit ID>",
+                "message": "<the commit message>",
+                <more fields, not important to this tutorial>
+            },
+            {
+                "id": "<the commit ID>",
+                "message": "<the commit message>",
+                <more fields, not important to this tutorial>
+            },
+            <more commits...>
+        ]
+    }
+
+In our program, the HTTP request is sent using the following
+function::
+
+    static char *request(const char *url);
+
+It takes the URL as a parameter, preforms a HTTP GET request, and
+returns a newly allocated string that contains the response body. If
+the request fails, an error message is printed to stderr and the
+return value is *NULL*. For full details, refer to :download:`the code
+<github_commits.c>`, as the actual implementation is not important
+here.
+
+.. _GitHub commits API: http://develop.github.com/p/commits.html
+
+.. _tutorial-the-program:
+
+The Program
+===========
+
+First the includes::
+
+    #include <string.h>
+    #include <jansson.h>
+
+Like all the programs using Jansson, we need to include
+:file:`jansson.h`.
+
+The following definitions are used to build the GitHub commits API
+request URL::
+
+   #define URL_FORMAT   "http://github.com/api/v2/json/commits/list/%s/%s/master"
+   #define URL_SIZE     256
+
+The following function is used when formatting the result to find the
+first newline in the commit message::
+
+    /* Return the offset of the first newline in text or the length of
+       text if there's no newline */
+    static int newline_offset(const char *text)
+    {
+        const char *newline = strchr(text, '\n');
+        if(!newline)
+            return strlen(text);
+        else
+            return (int)(newline - text);
+    }
+
+The main function follows. In the beginning, we first declare a bunch
+of variables and check the command line parameters::
+
+    unsigned int i;
+    char *text;
+    char url[URL_SIZE];
+
+    json_t *root;
+    json_error_t error;
+    json_t *commits;
+
+    if(argc != 3)
+    {
+        fprintf(stderr, "usage: %s USER REPOSITORY\n\n", argv[0]);
+        fprintf(stderr, "List commits at USER's REPOSITORY.\n\n");
+        return 2;
+    }
+
+Then we build the request URL using the user and repository names
+given as command line parameters::
+
+    snprintf(url, URL_SIZE, URL_FORMAT, argv[1], argv[2]);
+
+This uses the ``URL_SIZE`` and ``URL_FORMAT`` constants defined above.
+Now we're ready to actually request the JSON data over the web::
+
+    text = request(url);
+    if(!text)
+        return 1;
+
+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
+as a response::
+
+    root = json_loads(text, &error);
+    free(text);
+
+    if(!root)
+    {
+        fprintf(stderr, "error: on line %d: %s\n", error.line, error.text);
+        return 1;
+    }
+
+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`
+structure given as the second parameter. In this case, our program
+prints the error information out and returns 1 from the main function.
+
+Now we're ready to extract the data out of the decoded JSON response.
+The structure of the response JSON was explained in section
+:ref:`tutorial-github-commits-api`.
+
+First, we'll extract the ``commits`` array from the JSON response::
+
+    commits = json_object_get(root, "commits");
+    if(!json_is_array(commits))
+    {
+        fprintf(stderr, "error: commits is not an array\n");
+        return 1;
+    }
+
+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.
+
+Then we proceed to loop over all the commits in the array::
+
+    for(i = 0; i < json_array_size(commits); i++)
+    {
+        json_t *commit, *id, *message;
+        const char *message_text;
+
+        commit = json_array_get(commits, i);
+        if(!json_is_object(commit))
+        {
+            fprintf(stderr, "error: commit %d is not an object\n", i + 1);
+            return 1;
+        }
+    ...
+
+The function :cfunc:`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()`.
+We also check that the resulting value is a JSON object.
+
+Next we'll extract the commit ID and commit message, and check that
+they both are JSON strings::
+
+        id = json_object_get(commit, "id");
+        if(!json_is_string(id))
+        {
+            fprintf(stderr, "error: commit %d: id is not a string\n", i + 1);
+            return 1;
+        }
+
+        message = json_object_get(commit, "message");
+        if(!json_is_string(message))
+        {
+            fprintf(stderr, "error: commit %d: message is not a string\n", i + 1);
+            return 1;
+        }
+    ...
+
+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()`::
+
+        message_text = json_string_value(message);
+        printf("%.8s %.*s\n",
+               json_string_value(id),
+               newline_offset(message_text),
+               message_text);
+    }
+
+After sending the HTTP request, we decoded the JSON text using
+:cfunc:`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
+Jansson can release the resources::
+
+    json_decref(root);
+    return 0;
+
+For a detailed explanation of reference counting in Jansson, see
+:ref:`apiref-reference-count` in :ref:`apiref`.
+
+The program's ready, let's test it and view the latest commits in
+Jansson's repository::
+
+    $ ./github_commits akheron jansson
+    86dc1d62 Fix indentation
+    b67e130f json_dumpf: Document the output shortage on error
+    4cd77771 Enhance handling of circular references
+    79009e62 json_dumps: Close the strbuffer if dumping fails
+    76999799 doc: Fix a small typo in apiref
+    22af193a doc/Makefile.am: Remove *.pyc in clean
+    951d091f Make integer, real and string mutable
+    185e107d Don't use non-portable asprintf()
+    ca7703fb Merge branch '1.0'
+    12cd4e8c jansson 1.0.4
+    <etc...>
+
+
+Conclusion
+==========
+
+In this tutorial, we implemented a program that fetches the latest
+commits of a GitHub repository using the GitHub commits API. Jansson
+was used to decode the JSON response and to extract the commit data.
+
+This tutorial only covered a small part of Jansson. For example, we
+did not create or manipulate JSON values at all. Proceed to
+:ref:`apiref` to explore all features of Jansson.

jansson.pc.in

@@ -0,0 +1,10 @@
+prefix=@prefix@
+exec_prefix=@exec_prefix@
+libdir=@libdir@
+includedir=${prefix}/include
+
+Name: Jansson
+Description: Library for encoding, decoding and manipulating JSON data
+Version: @VERSION@
+Libs: -L${libdir} -ljansson
+Cflags: -I${includedir}

src/Makefile.am

@@ -13,6 +13,8 @@ libjansson_la_SOURCES = \
 	utf.h \
 	util.h \
 	value.c
-libjansson_la_LDFLAGS = -version-info 0:4:0
+libjansson_la_LDFLAGS = \
+	-export-symbols-regex '^json_' \
+	-version-info 3:0:3
 
-AM_CFLAGS = -Wall -Wextra -Werror -std=c99
+AM_CFLAGS = -Wall -Wextra -Werror

src/dump.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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.
@@ -9,10 +9,15 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#include <unistd.h>
+#include <assert.h>
 
 #include <jansson.h>
+#include "jansson_private.h"
 #include "strbuffer.h"
+#include "utf.h"
+
+#define MAX_INTEGER_STR_LENGTH  100
+#define MAX_REAL_STR_LENGTH     100
 
 typedef int (*dump_func)(const char *buffer, int size, void *data);
 
@@ -36,54 +41,74 @@ static int dump_to_file(const char *buffer, int size, void *data)
     return 0;
 }
 
-static int dump_indent(uint32_t flags, int depth, dump_func dump, void *data)
+/* 256 spaces (the maximum indentation size) */
+static char whitespace[] = "                                                                                                                                                                                                                                                                ";
+
+static int dump_indent(unsigned long flags, int depth, int space, dump_func dump, void *data)
 {
     if(JSON_INDENT(flags) > 0)
     {
-        char *ws_buffer;
-        int ws_count = JSON_INDENT(flags) * depth;
+        int i, ws_count = JSON_INDENT(flags);
 
         if(dump("\n", 1, data))
             return -1;
 
-        if(ws_count == 0)
-            return 0;
-
-        ws_buffer = alloca(ws_count);
-        memset(ws_buffer, ' ', ws_count);
-        return dump(ws_buffer, ws_count, data);
+        for(i = 0; i < depth; i++)
+        {
+            if(dump(whitespace, ws_count, data))
+                return -1;
+        }
+    }
+    else if(space && !(flags & JSON_COMPACT))
+    {
+        return dump(" ", 1, data);
     }
     return 0;
 }
 
-static int dump_string(const char *str, dump_func dump, void *data)
+static int dump_string(const char *str, int ascii, dump_func dump, void *data)
 {
-    const char *end;
+    const char *pos, *end;
+    int32_t codepoint;
 
     if(dump("\"", 1, data))
         return -1;
 
-    end = str;
+    end = pos = str;
     while(1)
     {
         const char *text;
-        char seq[7];
+        char seq[13];
         int length;
 
-        while(*end && *end != '\\' && *end != '"' && (unsigned char)*end > 0x1F)
-            end++;
+        while(*end)
+        {
+            end = utf8_iterate(pos, &codepoint);
+            if(!end)
+                return -1;
 
-        if(end != str) {
-            if(dump(str, end - str, data))
+            /* mandatory escape or control char */
+            if(codepoint == '\\' || codepoint == '"' || codepoint < 0x20)
+                break;
+
+            /* non-ASCII */
+            if(ascii && codepoint > 0x7F)
+                break;
+
+            pos = end;
+        }
+
+        if(pos != str) {
+            if(dump(str, pos - str, data))
                 return -1;
         }
 
-        if(!*end)
+        if(end == pos)
             break;
 
         /* handle \, ", and control codes */
         length = 2;
-        switch(*end)
+        switch(codepoint)
         {
             case '\\': text = "\\\\"; break;
             case '\"': text = "\\\""; break;
@@ -94,9 +119,27 @@ static int dump_string(const char *str, dump_func dump, void *data)
             case '\t': text = "\\t"; break;
             default:
             {
-                sprintf(seq, "\\u00%02x", *end);
+                /* codepoint is in BMP */
+                if(codepoint < 0x10000)
+                {
+                    sprintf(seq, "\\u%04x", codepoint);
+                    length = 6;
+                }
+
+                /* not in BMP -> construct a UTF-16 surrogate pair */
+                else
+                {
+                    int32_t first, last;
+
+                    codepoint -= 0x10000;
+                    first = 0xD800 | ((codepoint & 0xffc00) >> 10);
+                    last = 0xDC00 | (codepoint & 0x003ff);
+
+                    sprintf(seq, "\\u%04x\\u%04x", first, last);
+                    length = 12;
+                }
+
                 text = seq;
-                length = 6;
                 break;
             }
         }
@@ -104,16 +147,29 @@ static int dump_string(const char *str, dump_func dump, void *data)
         if(dump(text, length, data))
             return -1;
 
-        end++;
-        str = end;
+        str = pos = end;
     }
 
     return dump("\"", 1, data);
 }
 
-static int do_dump(const json_t *json, uint32_t flags, int depth,
+static int object_key_compare_keys(const void *key1, const void *key2)
+{
+    return strcmp((*(const object_key_t **)key1)->key,
+                  (*(const object_key_t **)key2)->key);
+}
+
+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, unsigned long flags, int depth,
                    dump_func dump, void *data)
 {
+    int ascii = flags & JSON_ENSURE_ASCII ? 1 : 0;
+
     switch(json_typeof(json)) {
         case JSON_NULL:
             return dump("null", 4, data);
@@ -126,103 +182,230 @@ static int do_dump(const json_t *json, uint32_t flags, int depth,
 
         case JSON_INTEGER:
         {
-            char *buffer;
-            int size, ret;
+            char buffer[MAX_INTEGER_STR_LENGTH];
+            int size;
 
-            size = asprintf(&buffer, "%d", json_integer_value(json));
-            if(size == -1)
+            size = snprintf(buffer, MAX_INTEGER_STR_LENGTH, "%d", json_integer_value(json));
+            if(size >= MAX_INTEGER_STR_LENGTH)
                 return -1;
 
-            ret = dump(buffer, size, data);
-            free(buffer);
-            return ret;
+            return dump(buffer, size, data);
         }
 
         case JSON_REAL:
         {
-            char *buffer;
-            int size, ret;
+            char buffer[MAX_REAL_STR_LENGTH];
+            int size;
 
-            size = asprintf(&buffer, "%.17f", json_real_value(json));
-            if(size == -1)
+            size = snprintf(buffer, MAX_REAL_STR_LENGTH, "%.17g",
+                            json_real_value(json));
+            if(size >= MAX_REAL_STR_LENGTH)
                 return -1;
 
-            ret = dump(buffer, size, data);
-            free(buffer);
-            return ret;
+            /* 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);
         }
 
         case JSON_STRING:
-            return dump_string(json_string_value(json), dump, data);
+            return dump_string(json_string_value(json), ascii, dump, data);
 
         case JSON_ARRAY:
         {
             int i;
-            int n = json_array_size(json);
+            int n;
+            json_array_t *array;
+
+            /* detect circular references */
+            array = json_to_array(json);
+            if(array->visited)
+                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, dump, data))
-                return -1;
+            }
+            if(dump_indent(flags, depth + 1, 0, dump, data))
+                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, dump, data))
-                        return -1;
+                       dump_indent(flags, depth + 1, 1, dump, data))
+                        goto array_error;
                 }
                 else
                 {
-                    if(dump_indent(flags, depth, dump, data))
-                        return -1;
+                    if(dump_indent(flags, depth, 0, dump, data))
+                        goto array_error;
                 }
             }
+
+            array->visited = 0;
             return dump("]", 1, data);
+
+        array_error:
+            array->visited = 0;
+            return -1;
         }
 
         case JSON_OBJECT:
         {
-            void *iter = json_object_iter((json_t *)json);
+            json_object_t *object;
+            void *iter;
+            const char *separator;
+            int separator_length;
+
+            if(flags & JSON_COMPACT) {
+                separator = ":";
+                separator_length = 1;
+            }
+            else {
+                separator = ": ";
+                separator_length = 2;
+            }
+
+            /* detect circular references */
+            object = json_to_object(json);
+            if(object->visited)
+                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, dump, data))
-                return -1;
-
-            while(iter)
-            {
-                void *next = json_object_iter_next((json_t *)json, iter);
-
-                dump_string(json_object_iter_key(iter), dump, data);
-                if(dump(": ", 2, data) ||
-                   do_dump(json_object_iter_value(iter), flags, depth + 1,
-                           dump, data))
-                    return -1;
-
-                if(next)
-                {
-                    if(dump(",", 1, data) ||
-                       dump_indent(flags, depth + 1, dump, data))
-                        return -1;
-                }
-                else
-                {
-                    if(dump_indent(flags, depth, dump, data))
-                        return -1;
-                }
-
-                iter = next;
             }
+            if(dump_indent(flags, depth + 1, 0, dump, data))
+                goto object_error;
+
+            if(flags & JSON_SORT_KEYS || flags & JSON_PRESERVE_ORDER)
+            {
+                const object_key_t **keys;
+                unsigned int size;
+                unsigned int i;
+                int (*cmp_func)(const void *, const void *);
+
+                size = json_object_size(json);
+                keys = malloc(size * sizeof(object_key_t *));
+                if(!keys)
+                    goto object_error;
+
+                i = 0;
+                while(iter)
+                {
+                    keys[i] = jsonp_object_iter_fullkey(iter);
+                    iter = json_object_iter_next((json_t *)json, iter);
+                    i++;
+                }
+                assert(i == size);
+
+                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;
+                    value = json_object_get(json, key);
+                    assert(value);
+
+                    dump_string(key, ascii, dump, data);
+                    if(dump(separator, separator_length, data) ||
+                       do_dump(value, flags, depth + 1, dump, data))
+                    {
+                        free(keys);
+                        goto object_error;
+                    }
+
+                    if(i < size - 1)
+                    {
+                        if(dump(",", 1, data) ||
+                           dump_indent(flags, depth + 1, 1, dump, data))
+                        {
+                            free(keys);
+                            goto object_error;
+                        }
+                    }
+                    else
+                    {
+                        if(dump_indent(flags, depth, 0, dump, data))
+                        {
+                            free(keys);
+                            goto object_error;
+                        }
+                    }
+                }
+
+                free(keys);
+            }
+            else
+            {
+                /* Don't sort keys */
+
+                while(iter)
+                {
+                    void *next = json_object_iter_next((json_t *)json, iter);
+
+                    dump_string(json_object_iter_key(iter), ascii, dump, data);
+                    if(dump(separator, separator_length, data) ||
+                       do_dump(json_object_iter_value(iter), flags, depth + 1,
+                               dump, data))
+                        goto object_error;
+
+                    if(next)
+                    {
+                        if(dump(",", 1, data) ||
+                           dump_indent(flags, depth + 1, 1, dump, data))
+                            goto object_error;
+                    }
+                    else
+                    {
+                        if(dump_indent(flags, depth, 0, dump, data))
+                            goto object_error;
+                    }
+
+                    iter = next;
+                }
+            }
+
+            object->visited = 0;
             return dump("}", 1, data);
+
+        object_error:
+            object->visited = 0;
+            return -1;
         }
 
         default:
@@ -232,7 +415,7 @@ static int do_dump(const json_t *json, uint32_t flags, int depth,
 }
 
 
-char *json_dumps(const json_t *json, uint32_t flags)
+char *json_dumps(const json_t *json, unsigned long flags)
 {
     strbuffer_t strbuff;
     char *result;
@@ -241,13 +424,12 @@ char *json_dumps(const json_t *json, uint32_t flags)
         return NULL;
 
     if(strbuffer_init(&strbuff))
-      return NULL;
-
-    if(do_dump(json, flags, 0, dump_to_strbuffer, (void *)&strbuff))
         return NULL;
 
-    if(dump_to_strbuffer("\n", 1, (void *)&strbuff))
+    if(do_dump(json, flags, 0, dump_to_strbuffer, (void *)&strbuff)) {
+        strbuffer_close(&strbuff);
         return NULL;
+    }
 
     result = strdup(strbuffer_value(&strbuff));
     strbuffer_close(&strbuff);
@@ -255,17 +437,15 @@ char *json_dumps(const json_t *json, uint32_t flags)
     return result;
 }
 
-int json_dumpf(const json_t *json, FILE *output, uint32_t flags)
+int json_dumpf(const json_t *json, FILE *output, unsigned long flags)
 {
     if(!json_is_array(json) && !json_is_object(json))
         return -1;
 
-    if(do_dump(json, flags, 0, dump_to_file, (void *)output))
-        return -1;
-    return dump_to_file("\n", 1, (void *)output);
+    return do_dump(json, flags, 0, dump_to_file, (void *)output);
 }
 
-int json_dump_file(const json_t *json, const char *path, uint32_t flags)
+int json_dump_file(const json_t *json, const char *path, unsigned long flags)
 {
     int result;
 

src/hashtable.c

@@ -1,10 +1,12 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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 <config.h>
+
 #include <stdlib.h>
 #include "hashtable.h"
 
@@ -133,6 +135,23 @@ static int hashtable_do_del(hashtable_t *hashtable,
     return 0;
 }
 
+static void hashtable_do_clear(hashtable_t *hashtable)
+{
+    list_t *list, *next;
+    pair_t *pair;
+
+    for(list = hashtable->list.next; list != &hashtable->list; list = next)
+    {
+        next = list->next;
+        pair = list_to_pair(list);
+        if(hashtable->free_key)
+            hashtable->free_key(pair->key);
+        if(hashtable->free_value)
+            hashtable->free_value(pair->value);
+        free(pair);
+    }
+}
+
 static int hashtable_do_rehash(hashtable_t *hashtable)
 {
     list_t *list, *next;
@@ -220,19 +239,7 @@ int hashtable_init(hashtable_t *hashtable,
 
 void hashtable_close(hashtable_t *hashtable)
 {
-    list_t *list, *next;
-    pair_t *pair;
-    for(list = hashtable->list.next; list != &hashtable->list; list = next)
-    {
-        next = list->next;
-        pair = list_to_pair(list);
-        if(hashtable->free_key)
-            hashtable->free_key(pair->key);
-        if(hashtable->free_value)
-            hashtable->free_value(pair->value);
-        free(pair);
-    }
-
+    hashtable_do_clear(hashtable);
     free(hashtable->buckets);
 }
 
@@ -242,31 +249,39 @@ int hashtable_set(hashtable_t *hashtable, void *key, void *value)
     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);
-
     /* 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 = 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;
 }
 
@@ -292,11 +307,43 @@ int hashtable_del(hashtable_t *hashtable, const void *key)
     return hashtable_do_del(hashtable, key, hash);
 }
 
+void hashtable_clear(hashtable_t *hashtable)
+{
+    unsigned int i;
+
+    hashtable_do_clear(hashtable);
+
+    for(i = 0; i < num_buckets(hashtable); i++)
+    {
+        hashtable->buckets[i].first = hashtable->buckets[i].last =
+            &hashtable->list;
+    }
+
+    list_init(&hashtable->list);
+    hashtable->size = 0;
+}
+
 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;
+    unsigned int 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;
@@ -316,3 +363,13 @@ 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, 2010 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.
@@ -13,8 +13,8 @@ typedef int (*key_cmp_fn)(const void *key1, const void *key2);
 typedef void (*free_fn)(void *key);
 
 struct hashtable_list {
-  struct hashtable_list *prev;
-  struct hashtable_list *next;
+    struct hashtable_list *prev;
+    struct hashtable_list *next;
 };
 
 struct hashtable_pair {
@@ -134,6 +134,15 @@ void *hashtable_get(hashtable_t *hashtable, const void *key);
  */
 int hashtable_del(hashtable_t *hashtable, const void *key);
 
+/**
+ * hashtable_clear - Clear hashtable
+ *
+ * @hashtable: The hashtable object
+ *
+ * Removes all items from the hashtable.
+ */
+void hashtable_clear(hashtable_t *hashtable);
+
 /**
  * hashtable_iter - Iterate over hashtable
  *
@@ -151,6 +160,17 @@ int hashtable_del(hashtable_t *hashtable, const void *key);
  */
 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
  *
@@ -176,4 +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

@@ -1,112 +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 JANSSON_H
-#define JANSSON_H
-
-#include <stdio.h>
-#include <stdint.h>
-
-/* types */
-
-typedef enum {
-    JSON_OBJECT,
-    JSON_ARRAY,
-    JSON_STRING,
-    JSON_INTEGER,
-    JSON_REAL,
-    JSON_TRUE,
-    JSON_FALSE,
-    JSON_NULL
-} json_type;
-
-typedef struct {
-    json_type type;
-    unsigned long refcount;
-} json_t;
-
-#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)
-#define json_is_string(json)   (json && json_typeof(json) == JSON_STRING)
-#define json_is_integer(json)  (json && json_typeof(json) == JSON_INTEGER)
-#define json_is_real(json)     (json && json_typeof(json) == JSON_REAL)
-#define json_is_number(json)   (json_is_integer(json) || json_is_real(json))
-#define json_is_true(json)     (json && json_typeof(json) == JSON_TRUE)
-#define json_is_false(json)    (json && json_typeof(json) == JSON_FALSE)
-#define json_is_boolean(json)  (json_is_true(json) || json_is_false(json))
-#define json_is_null(json)     (json && json_typeof(json) == JSON_NULL)
-
-/* construction, destruction, reference counting */
-
-json_t *json_object(void);
-json_t *json_array(void);
-json_t *json_string(const char *value);
-json_t *json_integer(int 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)
-{
-    if(json)
-        ++json->refcount;
-    return json;
-}
-
-/* do not call json_delete directly */
-void json_delete(json_t *json);
-
-static inline void json_decref(json_t *json)
-{
-    if(json && --json->refcount == 0)
-        json_delete(json);
-}
-
-
-/* getters, setters, manipulation */
-
-json_t *json_object_get(const json_t *object, const char *key);
-int json_object_set(json_t *object, const char *key, json_t *value);
-int json_object_del(json_t *object, const char *key);
-void *json_object_iter(json_t *object);
-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);
-
-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(json_t *array, unsigned int index, json_t *value);
-int json_array_append(json_t *array, json_t *value);
-
-const char *json_string_value(const json_t *json);
-int json_integer_value(const json_t *json);
-double json_real_value(const json_t *json);
-double json_number_value(const json_t *json);
-
-
-/* loading, printing */
-
-#define JSON_ERROR_TEXT_LENGTH  160
-
-typedef struct {
-    char text[JSON_ERROR_TEXT_LENGTH];
-    int line;
-} json_error_t;
-
-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)
-
-char *json_dumps(const json_t *json, uint32_t flags);
-int json_dumpf(const json_t *json, FILE *output, uint32_t flags);
-int json_dump_file(const json_t *json, const char *path, uint32_t flags);
-
-#endif

src/jansson.h.in

@@ -0,0 +1,191 @@
+/*
+ * Copyright (c) 2009, 2010 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 JANSSON_H
+#define JANSSON_H
+
+#include <stdio.h>
+
+#ifndef __cplusplus
+#define JSON_INLINE @json_inline@
+#else
+#define JSON_INLINE inline
+extern "C" {
+#endif
+
+/* types */
+
+typedef enum {
+    JSON_OBJECT,
+    JSON_ARRAY,
+    JSON_STRING,
+    JSON_INTEGER,
+    JSON_REAL,
+    JSON_TRUE,
+    JSON_FALSE,
+    JSON_NULL
+} json_type;
+
+typedef struct {
+    json_type type;
+    unsigned long refcount;
+} json_t;
+
+#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)
+#define json_is_string(json)   (json && json_typeof(json) == JSON_STRING)
+#define json_is_integer(json)  (json && json_typeof(json) == JSON_INTEGER)
+#define json_is_real(json)     (json && json_typeof(json) == JSON_REAL)
+#define json_is_number(json)   (json_is_integer(json) || json_is_real(json))
+#define json_is_true(json)     (json && json_typeof(json) == JSON_TRUE)
+#define json_is_false(json)    (json && json_typeof(json) == JSON_FALSE)
+#define json_is_boolean(json)  (json_is_true(json) || json_is_false(json))
+#define json_is_null(json)     (json && json_typeof(json) == JSON_NULL)
+
+/* construction, destruction, reference counting */
+
+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_real(double value);
+json_t *json_true(void);
+json_t *json_false(void);
+json_t *json_null(void);
+
+static JSON_INLINE
+json_t *json_incref(json_t *json)
+{
+    if(json && json->refcount != (unsigned int)-1)
+        ++json->refcount;
+    return json;
+}
+
+/* do not call json_delete directly */
+void json_delete(json_t *json);
+
+static JSON_INLINE
+void json_decref(json_t *json)
+{
+    if(json && json->refcount != (unsigned int)-1 && --json->refcount == 0)
+        json_delete(json);
+}
+
+
+/* getters, setters, manipulation */
+
+unsigned int 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);
+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 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 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));
+}
+
+static 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));
+}
+
+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);
+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_clear(json_t *array);
+int json_array_extend(json_t *array, json_t *other);
+
+static JSON_INLINE
+int json_array_set(json_t *array, unsigned int index, json_t *value)
+{
+    return json_array_set_new(array, index, json_incref(value));
+}
+
+static JSON_INLINE
+int json_array_append(json_t *array, json_t *value)
+{
+    return json_array_append_new(array, json_incref(value));
+}
+
+static JSON_INLINE
+int json_array_insert(json_t *array, unsigned int 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);
+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_real_set(json_t *real, double value);
+
+
+/* equality */
+
+int json_equal(json_t *value1, json_t *value2);
+
+
+/* copying */
+
+json_t *json_copy(json_t *value);
+json_t *json_deep_copy(json_t *value);
+
+
+/* loading, printing */
+
+#define JSON_ERROR_TEXT_LENGTH  160
+
+typedef struct {
+    char text[JSON_ERROR_TEXT_LENGTH];
+    int line;
+} json_error_t;
+
+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
+#define JSON_PRESERVE_ORDER 0x800
+
+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);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

src/jansson_private.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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.
@@ -8,8 +8,53 @@
 #ifndef JANSSON_PRIVATE_H
 #define JANSSON_PRIVATE_H
 
-int json_object_set_nocheck(json_t *json, const char *key, json_t *value);
-json_t *json_string_nocheck(const char *value);
+#include "jansson.h"
+#include "hashtable.h"
 
+#define container_of(ptr_, type_, member_)  \
+    ((type_ *)((char *)ptr_ - (size_t)&((type_ *)0)->member_))
+
+typedef struct {
+    json_t json;
+    hashtable_t hashtable;
+    unsigned long serial;
+    int visited;
+} json_object_t;
+
+typedef struct {
+    json_t json;
+    unsigned int size;
+    unsigned int entries;
+    json_t **table;
+    int visited;
+} json_array_t;
+
+typedef struct {
+    json_t json;
+    char *value;
+} json_string_t;
+
+typedef struct {
+    json_t json;
+    double value;
+} json_real_t;
+
+typedef struct {
+    json_t json;
+    int value;
+} json_integer_t;
+
+#define json_to_object(json_)  container_of(json_, json_object_t, json)
+#define json_to_array(json_)   container_of(json_, json_array_t, json)
+#define json_to_string(json_)  container_of(json_, json_string_t, json)
+#define json_to_real(json_)   container_of(json_, json_real_t, json)
+#define json_to_integer(json_) container_of(json_, json_integer_t, json)
+
+typedef struct {
+    unsigned long serial;
+    char key[];
+} object_key_t;
+
+const object_key_t *jsonp_object_iter_fullkey(void *iter);
 
 #endif

src/load.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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.
@@ -13,7 +13,6 @@
 #include <stdlib.h>
 #include <string.h>
 #include <stdarg.h>
-#include <unistd.h>
 #include <assert.h>
 
 #include <jansson.h>
@@ -114,7 +113,8 @@ static void error_set(json_error_t *error, const lex_t *lex,
 
 /*** 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, eof_func eof, void *data)
 {
     stream->get = get;
     stream->eof = eof;
@@ -149,7 +149,7 @@ static char stream_get(stream_t *stream, json_error_t *error)
             for(i = 1; i < count; i++)
                 stream->buffer[i] = stream->get(stream->data);
 
-            if(!utf8_check_full(stream->buffer, count))
+            if(!utf8_check_full(stream->buffer, count, NULL))
                 goto out;
 
             stream->stream_pos += count;
@@ -222,10 +222,10 @@ static void lex_save_cached(lex_t *lex)
 }
 
 /* assumes that str points to 'u' plus at least 4 valid hex digits */
-static int decode_unicode_escape(const char *str)
+static int32_t decode_unicode_escape(const char *str)
 {
     int i;
-    int value = 0;
+    int32_t value = 0;
 
     assert(str[0] == 'u');
 
@@ -326,7 +326,7 @@ static void lex_scan_string(lex_t *lex, json_error_t *error)
             if(*p == 'u') {
                 char buffer[4];
                 int length;
-                int value;
+                int32_t value;
 
                 value = decode_unicode_escape(p);
                 p += 5;
@@ -334,14 +334,15 @@ static void lex_scan_string(lex_t *lex, json_error_t *error)
                 if(0xD800 <= value && value <= 0xDBFF) {
                     /* surrogate pair */
                     if(*p == '\\' && *(p + 1) == 'u') {
-                        int value2 = decode_unicode_escape(++p);
+                        int32_t value2 = decode_unicode_escape(++p);
                         p += 5;
 
                         if(0xDC00 <= value2 && value2 <= 0xDFFF) {
                             /* valid second surrogate */
-                            value = ((value - 0xD800) << 10) +
-                                    (value2 - 0xDC00) +
-                                    0x10000;
+                            value =
+                                ((value - 0xD800) << 10) +
+                                (value2 - 0xDC00) +
+                                0x10000;
                         }
                         else {
                             /* invalid second surrogate */
@@ -418,11 +419,15 @@ static int lex_scan_number(lex_t *lex, char c, json_error_t *error)
             goto out;
         }
     }
-    else /* c != '0' */ {
+    else if(isdigit(c)) {
         c = lex_get_save(lex, error);
         while(isdigit(c))
             c = lex_get_save(lex, error);
     }
+    else {
+      lex_unget_unsave(lex, c);
+      goto out;
+    }
 
     if(c != '.' && c != 'E' && c != 'e') {
         long value;
@@ -479,14 +484,7 @@ static int lex_scan_number(lex_t *lex, char c, json_error_t *error)
     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(errno == ERANGE && value != 0) {
         error_set(error, lex, "real number overflow");
         goto out;
     }
@@ -506,8 +504,8 @@ static int lex_scan(lex_t *lex, json_error_t *error)
     strbuffer_clear(&lex->saved_text);
 
     if(lex->token == TOKEN_STRING) {
-      free(lex->value.string);
-      lex->value.string = NULL;
+        free(lex->value.string);
+        lex->value.string = NULL;
     }
 
     c = lex_get(lex, error);
@@ -572,6 +570,17 @@ out:
     return lex->token;
 }
 
+static char *lex_steal_string(lex_t *lex)
+{
+    char *result = NULL;
+    if(lex->token == TOKEN_STRING)
+    {
+        result = lex->value.string;
+        lex->value.string = NULL;
+    }
+    return result;
+}
+
 static int lex_init(lex_t *lex, get_func get, eof_func eof, void *data)
 {
     stream_init(&lex->stream, get, eof, data);
@@ -615,7 +624,7 @@ static json_t *parse_object(lex_t *lex, json_error_t *error)
             goto error;
         }
 
-        key = strdup(lex->value.string);
+        key = lex_steal_string(lex);
         if(!key)
             return NULL;
 
@@ -734,7 +743,7 @@ static json_t *parse_value(lex_t *lex, json_error_t *error)
             break;
 
         case '{':
-          json = parse_object(lex, error);
+            json = parse_object(lex, error);
             break;
 
         case '[':
@@ -756,7 +765,7 @@ static json_t *parse_value(lex_t *lex, json_error_t *error)
     return json;
 }
 
-json_t *parse_json(lex_t *lex, json_error_t *error)
+static json_t *parse_json(lex_t *lex, json_error_t *error)
 {
     error_init(error);
 
@@ -853,6 +862,8 @@ json_t *json_load_file(const char *path, json_error_t *error)
     json_t *result;
     FILE *fp;
 
+    error_init(error);
+
     fp = fopen(path, "r");
     if(!fp)
     {

src/strbuffer.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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.

src/strbuffer.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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.

src/utf.c

@@ -1,13 +1,14 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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 "utf.h"
 
-int utf8_encode(int codepoint, char *buffer, int *size)
+int utf8_encode(int32_t codepoint, char *buffer, int *size)
 {
     if(codepoint < 0)
         return -1;
@@ -79,9 +80,10 @@ int utf8_check_first(char byte)
     }
 }
 
-int utf8_check_full(const char *buffer, int size)
+int utf8_check_full(const char *buffer, int size, int32_t *codepoint)
 {
-    int i, value = 0;
+    int i;
+    int32_t value = 0;
     unsigned char u = (unsigned char)buffer[0];
 
     if(size == 2)
@@ -128,9 +130,38 @@ int utf8_check_full(const char *buffer, int size)
         return 0;
     }
 
+    if(codepoint)
+        *codepoint = value;
+
     return 1;
 }
 
+const char *utf8_iterate(const char *buffer, int32_t *codepoint)
+{
+    int count;
+    int32_t value;
+
+    if(!*buffer)
+        return buffer;
+
+    count = utf8_check_first(buffer[0]);
+    if(count <= 0)
+        return NULL;
+
+    if(count == 1)
+        value = (unsigned char)buffer[0];
+    else
+    {
+        if(!utf8_check_full(buffer, count, &value))
+            return NULL;
+    }
+
+    if(codepoint)
+        *codepoint = value;
+
+    return buffer + count;
+}
+
 int utf8_check_string(const char *string, int length)
 {
     int i;
@@ -148,7 +179,7 @@ int utf8_check_string(const char *string, int length)
             if(i + count > length)
                 return 0;
 
-            if(!utf8_check_full(&string[i], count))
+            if(!utf8_check_full(&string[i], count, NULL))
                 return 0;
 
             i += count - 1;

src/utf.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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.
@@ -8,10 +8,20 @@
 #ifndef UTF_H
 #define UTF_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
+
 int utf8_encode(int codepoint, char *buffer, int *size);
 
 int utf8_check_first(char byte);
-int utf8_check_full(const char *buffer, int size);
+int utf8_check_full(const char *buffer, int size, int32_t *codepoint);
+const char *utf8_iterate(const char *buffer, int32_t *codepoint);
 
 int utf8_check_string(const char *string, int length);
 

src/util.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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.

src/value.c

@@ -1,11 +1,14 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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 <config.h>
+
 #include <stdlib.h>
 #include <string.h>
 
@@ -15,41 +18,6 @@
 #include "utf.h"
 #include "util.h"
 
-#define container_of(ptr_, type_, member_)  \
-    ((type_ *)((char *)ptr_ - (size_t)&((type_ *)0)->member_))
-
-typedef struct {
-    json_t json;
-    hashtable_t hashtable;
-} json_object_t;
-
-typedef struct {
-    json_t json;
-    unsigned int size;
-    unsigned int entries;
-    json_t **table;
-} json_array_t;
-
-typedef struct {
-    json_t json;
-    char *value;
-} json_string_t;
-
-typedef struct {
-    json_t json;
-    double value;
-} json_real_t;
-
-typedef struct {
-    json_t json;
-    int value;
-} json_integer_t;
-
-#define json_to_object(json_)  container_of(json_, json_object_t, json)
-#define json_to_array(json_)   container_of(json_, json_array_t, json)
-#define json_to_string(json_)  container_of(json_, json_string_t, json)
-#define json_to_real(json_)   container_of(json_, json_real_t, json)
-#define json_to_integer(json_) container_of(json_, json_integer_t, json)
 
 static inline void json_init(json_t *json, json_type type)
 {
@@ -60,9 +28,16 @@ static inline void json_init(json_t *json, json_type type)
 
 /*** object ***/
 
-static unsigned int hash_string(const void *key)
+/* 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 unsigned int hash_key(const void *ptr)
 {
-    const char *str = (const char *)key;
+    const char *str = ((const object_key_t *)ptr)->key;
+
     unsigned int hash = 5381;
     unsigned int c;
 
@@ -75,9 +50,10 @@ static unsigned int hash_string(const void *key)
     return hash;
 }
 
-static int string_equal(const void *key1, const void *key2)
+static int key_equal(const void *ptr1, const void *ptr2)
 {
-    return strcmp((const char *)key1, (const char *)key2) == 0;
+    return strcmp(((const object_key_t *)ptr1)->key,
+                  ((const object_key_t *)ptr2)->key) == 0;
 }
 
 static void value_decref(void *value)
@@ -92,12 +68,16 @@ json_t *json_object(void)
         return NULL;
     json_init(&object->json, JSON_OBJECT);
 
-    if(hashtable_init(&object->hashtable, hash_string, string_equal,
+    if(hashtable_init(&object->hashtable, hash_key, key_equal,
                       free, value_decref))
     {
         free(object);
         return NULL;
     }
+
+    object->serial = 0;
+    object->visited = 0;
+
     return &object->json;
 }
 
@@ -107,6 +87,17 @@ static void json_delete_object(json_object_t *object)
     free(object);
 }
 
+unsigned int json_object_size(const json_t *json)
+{
+    json_object_t *object;
+
+    if(!json_is_object(json))
+        return -1;
+
+    object = json_to_object(json);
+    return object->hashtable.size;
+}
+
 json_t *json_object_get(const json_t *json, const char *key)
 {
     json_object_t *object;
@@ -115,26 +106,49 @@ 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_nocheck(json_t *json, const char *key, json_t *value)
+int json_object_set_new_nocheck(json_t *json, const char *key, json_t *value)
 {
     json_object_t *object;
+    object_key_t *k;
 
-    if(!json_is_object(json))
+    if(!key || !value)
         return -1;
 
+    if(!json_is_object(json) || json == value)
+    {
+        json_decref(value);
+        return -1;
+    }
     object = json_to_object(json);
-    return hashtable_set(&object->hashtable, strdup(key), json_incref(value));
+
+    k = malloc(sizeof(object_key_t) + strlen(key) + 1);
+    if(!k)
+        return -1;
+
+    k->serial = object->serial++;
+    strcpy(k->key, key);
+
+    if(hashtable_set(&object->hashtable, k, value))
+    {
+        json_decref(value);
+        return -1;
+    }
+
+    return 0;
 }
 
-int json_object_set(json_t *json, const char *key, json_t *value)
+int json_object_set_new(json_t *json, const char *key, json_t *value)
 {
-    if(!utf8_check_string(key, -1))
+    if(!key || !utf8_check_string(key, -1))
+    {
+        json_decref(value);
         return -1;
+    }
 
-    return json_object_set_nocheck(json, key, value);
+    return json_object_set_new_nocheck(json, key, value);
 }
 
 int json_object_del(json_t *json, const char *key)
@@ -145,7 +159,44 @@ 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)
+{
+    json_object_t *object;
+
+    if(!json_is_object(json))
+        return -1;
+
+    object = json_to_object(json);
+    hashtable_clear(&object->hashtable);
+
+    return 0;
+}
+
+int json_object_update(json_t *object, json_t *other)
+{
+    void *iter;
+
+    if(!json_is_object(object) || !json_is_object(other))
+        return -1;
+
+    iter = json_object_iter(other);
+    while(iter) {
+        const char *key;
+        json_t *value;
+
+        key = json_object_iter_key(iter);
+        value = json_object_iter_value(iter);
+
+        if(json_object_set_nocheck(object, key, value))
+            return -1;
+
+        iter = json_object_iter_next(other, iter);
+    }
+
+    return 0;
 }
 
 void *json_object_iter(json_t *json)
@@ -159,6 +210,17 @@ 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;
@@ -170,12 +232,20 @@ 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)
@@ -186,6 +256,95 @@ 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;
+
+    if(json_object_size(object1) != json_object_size(object2))
+        return 0;
+
+    iter = json_object_iter(object1);
+    while(iter)
+    {
+        const char *key;
+        json_t *value1, *value2;
+
+        key = json_object_iter_key(iter);
+        value1 = json_object_iter_value(iter);
+        value2 = json_object_get(object2, key);
+
+        if(!json_equal(value1, value2))
+            return 0;
+
+        iter = json_object_iter_next(object1, iter);
+    }
+
+    return 1;
+}
+
+static json_t *json_object_copy(json_t *object)
+{
+    json_t *result;
+    void *iter;
+
+    result = json_object();
+    if(!result)
+        return NULL;
+
+    iter = json_object_iter(object);
+    while(iter)
+    {
+        const char *key;
+        json_t *value;
+
+        key = json_object_iter_key(iter);
+        value = json_object_iter_value(iter);
+        json_object_set_nocheck(result, key, value);
+
+        iter = json_object_iter_next(object, iter);
+    }
+
+    return result;
+}
+
+static json_t *json_object_deep_copy(json_t *object)
+{
+    json_t *result;
+    void *iter;
+
+    result = json_object();
+    if(!result)
+        return NULL;
+
+    iter = json_object_iter(object);
+    while(iter)
+    {
+        const char *key;
+        json_t *value;
+
+        key = json_object_iter_key(iter);
+        value = json_object_iter_value(iter);
+        json_object_set_new_nocheck(result, key, json_deep_copy(value));
+
+        iter = json_object_iter_next(object, iter);
+    }
+
+    return result;
+}
+
 
 /*** array ***/
 
@@ -193,12 +352,19 @@ json_t *json_array(void)
 {
     json_array_t *array = malloc(sizeof(json_array_t));
     if(!array)
-      return NULL;
+        return NULL;
     json_init(&array->json, JSON_ARRAY);
 
     array->entries = 0;
-    array->size = 0;
-    array->table = NULL;
+    array->size = 8;
+
+    array->table = malloc(array->size * sizeof(json_t *));
+    if(!array->table) {
+        free(array);
+        return NULL;
+    }
+
+    array->visited = 0;
 
     return &array->json;
 }
@@ -235,9 +401,143 @@ json_t *json_array_get(const json_t *json, unsigned int index)
     return array->table[index];
 }
 
-int json_array_set(json_t *json, unsigned int index, json_t *value)
+int json_array_set_new(json_t *json, unsigned int index, json_t *value)
 {
     json_array_t *array;
+
+    if(!value)
+        return -1;
+
+    if(!json_is_array(json) || json == value)
+    {
+        json_decref(value);
+        return -1;
+    }
+    array = json_to_array(json);
+
+    if(index >= array->entries)
+    {
+        json_decref(value);
+        return -1;
+    }
+
+    json_decref(array->table[index]);
+    array->table[index] = value;
+
+    return 0;
+}
+
+static void array_move(json_array_t *array, unsigned int dest,
+                       unsigned int src, unsigned int 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)
+{
+    memcpy(&dest[dpos], &src[spos], count * sizeof(json_t *));
+}
+
+static json_t **json_array_grow(json_array_t *array,
+                                unsigned int amount,
+                                int copy)
+{
+    unsigned int new_size;
+    json_t **old_table, **new_table;
+
+    if(array->entries + amount <= array->size)
+        return array->table;
+
+    old_table = array->table;
+
+    new_size = max(array->size + amount, array->size * 2);
+    new_table = malloc(new_size * sizeof(json_t *));
+    if(!new_table)
+        return NULL;
+
+    array->size = new_size;
+    array->table = new_table;
+
+    if(copy) {
+        array_copy(array->table, 0, old_table, 0, array->entries);
+        free(old_table);
+        return array->table;
+    }
+
+    return old_table;
+}
+
+int json_array_append_new(json_t *json, json_t *value)
+{
+    json_array_t *array;
+
+    if(!value)
+        return -1;
+
+    if(!json_is_array(json) || json == value)
+    {
+        json_decref(value);
+        return -1;
+    }
+    array = json_to_array(json);
+
+    if(!json_array_grow(array, 1, 1)) {
+        json_decref(value);
+        return -1;
+    }
+
+    array->table[array->entries] = value;
+    array->entries++;
+
+    return 0;
+}
+
+int json_array_insert_new(json_t *json, unsigned int index, json_t *value)
+{
+    json_array_t *array;
+    json_t **old_table;
+
+    if(!value)
+        return -1;
+
+    if(!json_is_array(json) || json == value) {
+        json_decref(value);
+        return -1;
+    }
+    array = json_to_array(json);
+
+    if(index > array->entries) {
+        json_decref(value);
+        return -1;
+    }
+
+    old_table = json_array_grow(array, 1, 0);
+    if(!old_table) {
+        json_decref(value);
+        return -1;
+    }
+
+    if(old_table != array->table) {
+        array_copy(array->table, 0, old_table, 0, index);
+        array_copy(array->table, index + 1, old_table, index,
+                   array->entries - index);
+        free(old_table);
+    }
+    else
+        array_move(array, index + 1, index, array->entries - index);
+
+    array->table[index] = value;
+    array->entries++;
+
+    return 0;
+}
+
+int json_array_remove(json_t *json, unsigned int index)
+{
+    json_array_t *array;
+
     if(!json_is_array(json))
         return -1;
     array = json_to_array(json);
@@ -246,48 +546,129 @@ int json_array_set(json_t *json, unsigned int index, json_t *value)
         return -1;
 
     json_decref(array->table[index]);
-    array->table[index] = json_incref(value);
+
+    array_move(array, index, index + 1, array->entries - index);
+    array->entries--;
 
     return 0;
 }
 
-int json_array_append(json_t *json, json_t *value)
+int json_array_clear(json_t *json)
 {
     json_array_t *array;
+    unsigned int i;
+
     if(!json_is_array(json))
         return -1;
     array = json_to_array(json);
 
-    if(array->entries == array->size) {
-        array->size = max(8, array->size * 2);
-        array->table = realloc(array->table, array->size * sizeof(json_t *));
-        if(!array->table)
-            return -1;
-    }
-
-    array->table[array->entries] = json_incref(value);
-    array->entries++;
+    for(i = 0; i < array->entries; i++)
+        json_decref(array->table[i]);
 
+    array->entries = 0;
     return 0;
 }
 
+int json_array_extend(json_t *json, json_t *other_json)
+{
+    json_array_t *array, *other;
+    unsigned int i;
+
+    if(!json_is_array(json) || !json_is_array(other_json))
+        return -1;
+    array = json_to_array(json);
+    other = json_to_array(other_json);
+
+    if(!json_array_grow(array, other->entries, 1))
+        return -1;
+
+    for(i = 0; i < other->entries; i++)
+        json_incref(other->table[i]);
+
+    array_copy(array->table, array->entries, other->table, 0, other->entries);
+
+    array->entries += other->entries;
+    return 0;
+}
+
+static int json_array_equal(json_t *array1, json_t *array2)
+{
+    unsigned int i, size;
+
+    size = json_array_size(array1);
+    if(size != json_array_size(array2))
+        return 0;
+
+    for(i = 0; i < size; i++)
+    {
+        json_t *value1, *value2;
+
+        value1 = json_array_get(array1, i);
+        value2 = json_array_get(array2, i);
+
+        if(!json_equal(value1, value2))
+            return 0;
+    }
+
+    return 1;
+}
+
+static json_t *json_array_copy(json_t *array)
+{
+    json_t *result;
+    unsigned int i;
+
+    result = json_array();
+    if(!result)
+        return NULL;
+
+    for(i = 0; i < json_array_size(array); i++)
+        json_array_append(result, json_array_get(array, i));
+
+    return result;
+}
+
+static json_t *json_array_deep_copy(json_t *array)
+{
+    json_t *result;
+    unsigned int i;
+
+    result = json_array();
+    if(!result)
+        return NULL;
+
+    for(i = 0; i < json_array_size(array); i++)
+        json_array_append_new(result, json_deep_copy(json_array_get(array, i)));
+
+    return result;
+}
 
 /*** string ***/
 
 json_t *json_string_nocheck(const char *value)
 {
-    json_string_t *string = malloc(sizeof(json_string_t));
+    json_string_t *string;
+
+    if(!value)
+        return NULL;
+
+    string = malloc(sizeof(json_string_t));
     if(!string)
-       return NULL;
+        return NULL;
     json_init(&string->json, JSON_STRING);
 
     string->value = strdup(value);
+    if(!string->value) {
+        free(string);
+        return NULL;
+    }
+
     return &string->json;
 }
 
 json_t *json_string(const char *value)
 {
-    if(!utf8_check_string(value, -1))
+    if(!value || !utf8_check_string(value, -1))
         return NULL;
 
     return json_string_nocheck(value);
@@ -301,12 +682,46 @@ const char *json_string_value(const json_t *json)
     return json_to_string(json)->value;
 }
 
+int json_string_set_nocheck(json_t *json, const char *value)
+{
+    char *dup;
+    json_string_t *string;
+
+    dup = strdup(value);
+    if(!dup)
+        return -1;
+
+    string = json_to_string(json);
+    free(string->value);
+    string->value = dup;
+
+    return 0;
+}
+
+int json_string_set(json_t *json, const char *value)
+{
+    if(!value || !utf8_check_string(value, -1))
+        return -1;
+
+    return json_string_set_nocheck(json, value);
+}
+
 static void json_delete_string(json_string_t *string)
 {
     free(string->value);
     free(string);
 }
 
+static int json_string_equal(json_t *string1, json_t *string2)
+{
+    return strcmp(json_string_value(string1), json_string_value(string2)) == 0;
+}
+
+static json_t *json_string_copy(json_t *string)
+{
+    return json_string_nocheck(json_string_value(string));
+}
+
 
 /*** integer ***/
 
@@ -314,7 +729,7 @@ json_t *json_integer(int value)
 {
     json_integer_t *integer = malloc(sizeof(json_integer_t));
     if(!integer)
-       return NULL;
+        return NULL;
     json_init(&integer->json, JSON_INTEGER);
 
     integer->value = value;
@@ -329,11 +744,31 @@ int json_integer_value(const json_t *json)
     return json_to_integer(json)->value;
 }
 
+int json_integer_set(json_t *json, int value)
+{
+    if(!json_is_integer(json))
+        return -1;
+
+    json_to_integer(json)->value = value;
+
+    return 0;
+}
+
 static void json_delete_integer(json_integer_t *integer)
 {
     free(integer);
 }
 
+static int json_integer_equal(json_t *integer1, json_t *integer2)
+{
+    return json_integer_value(integer1) == json_integer_value(integer2);
+}
+
+static json_t *json_integer_copy(json_t *integer)
+{
+    return json_integer(json_integer_value(integer));
+}
+
 
 /*** real ***/
 
@@ -341,7 +776,7 @@ json_t *json_real(double value)
 {
     json_real_t *real = malloc(sizeof(json_real_t));
     if(!real)
-       return NULL;
+        return NULL;
     json_init(&real->json, JSON_REAL);
 
     real->value = value;
@@ -356,11 +791,31 @@ double json_real_value(const json_t *json)
     return json_to_real(json)->value;
 }
 
-static void json_delete_real (json_real_t *real)
+int json_real_set(json_t *json, double value)
+{
+    if(!json_is_real(json))
+        return 0;
+
+    json_to_real(json)->value = value;
+
+    return 0;
+}
+
+static void json_delete_real(json_real_t *real)
 {
     free(real);
 }
 
+static int json_real_equal(json_t *real1, json_t *real2)
+{
+    return json_real_value(real1) == json_real_value(real2);
+}
+
+static json_t *json_real_copy(json_t *real)
+{
+    return json_real(json_real_value(real));
+}
+
 
 /*** number ***/
 
@@ -381,9 +836,9 @@ json_t *json_true(void)
 {
     static json_t the_true = {
         .type = JSON_TRUE,
-        .refcount = 1
+        .refcount = (unsigned int)-1
     };
-    return json_incref(&the_true);
+    return &the_true;
 }
 
 
@@ -391,9 +846,9 @@ json_t *json_false(void)
 {
     static json_t the_false = {
         .type = JSON_FALSE,
-        .refcount = 1
+        .refcount = (unsigned int)-1
     };
-    return json_incref(&the_false);
+    return &the_false;
 }
 
 
@@ -401,9 +856,9 @@ json_t *json_null(void)
 {
     static json_t the_null = {
         .type = JSON_NULL,
-        .refcount = 1
+        .refcount = (unsigned int)-1
     };
-    return json_incref(&the_null);
+    return &the_null;
 }
 
 
@@ -428,3 +883,94 @@ void json_delete(json_t *json)
 
     /* json_delete is not called for true, false or null */
 }
+
+
+/*** equality ***/
+
+int json_equal(json_t *json1, json_t *json2)
+{
+    if(!json1 || !json2)
+        return 0;
+
+    if(json_typeof(json1) != json_typeof(json2))
+        return 0;
+
+    /* this covers true, false and null as they are singletons */
+    if(json1 == json2)
+        return 1;
+
+    if(json_is_object(json1))
+        return json_object_equal(json1, json2);
+
+    if(json_is_array(json1))
+        return json_array_equal(json1, json2);
+
+    if(json_is_string(json1))
+        return json_string_equal(json1, json2);
+
+    if(json_is_integer(json1))
+        return json_integer_equal(json1, json2);
+
+    if(json_is_real(json1))
+        return json_real_equal(json1, json2);
+
+    return 0;
+}
+
+
+/*** copying ***/
+
+json_t *json_copy(json_t *json)
+{
+    if(!json)
+        return NULL;
+
+    if(json_is_object(json))
+        return json_object_copy(json);
+
+    if(json_is_array(json))
+        return json_array_copy(json);
+
+    if(json_is_string(json))
+        return json_string_copy(json);
+
+    if(json_is_integer(json))
+        return json_integer_copy(json);
+
+    if(json_is_real(json))
+        return json_real_copy(json);
+
+    if(json_is_true(json) || json_is_false(json) || json_is_null(json))
+        return json;
+
+    return NULL;
+}
+
+json_t *json_deep_copy(json_t *json)
+{
+    if(!json)
+        return NULL;
+
+    if(json_is_object(json))
+        return json_object_deep_copy(json);
+
+    if(json_is_array(json))
+        return json_array_deep_copy(json);
+
+    /* for the rest of the types, deep copying doesn't differ from
+       shallow copying */
+
+    if(json_is_string(json))
+        return json_string_copy(json);
+
+    if(json_is_integer(json))
+        return json_integer_copy(json);
+
+    if(json_is_real(json))
+        return json_real_copy(json);
+
+    if(json_is_true(json) || json_is_false(json) || json_is_null(json))
+        return json;
+
+    return NULL;
+}

test/.gitignore

@@ -1,7 +1,11 @@
-loadf_dumpf
-loads_dumps
-load_file_dump_file
-testlogs
-testprogs/test_array
-testprogs/test_number
-testprogs/test_object
+logs
+bin/json_process
+suites/api/test_array
+suites/api/test_equal
+suites/api/test_copy
+suites/api/test_dump
+suites/api/test_load
+suites/api/test_number
+suites/api/test_object
+suites/api/test_simple
+suites/api/test_cpp

test/Makefile.am

@@ -1,22 +1,10 @@
-DIST_SUBDIRS = testprogs testdata
-SUBDIRS = testprogs
+SUBDIRS = bin suites
+EXTRA_DIST = scripts run-suites
 
-check_PROGRAMS = loadf_dumpf loads_dumps load_file_dump_file
-
-AM_CPPFLAGS = -I$(top_srcdir)/src
-AM_CFLAGS = -Wall -Werror
-LDFLAGS = -static  # for speed and Valgrind
-LDADD = ../src/libjansson.la
-
-TESTS = test-api test-invalid test-valid
-
-EXTRA_DIST = \
-	test-api \
-	test-invalid \
-	test-valid \
-	run-test \
-	json-compare.py \
-	split-testfile.py
+TESTS = run-suites
+TESTS_ENVIRONMENT = \
+	top_srcdir=$(top_srcdir) \
+	top_builddir=$(top_builddir)
 
 clean-local:
-	rm -rf testlogs
+	rm -rf logs

test/bin/Makefile.am

@@ -0,0 +1,6 @@
+check_PROGRAMS = json_process
+
+AM_CPPFLAGS = -I$(top_srcdir)/src
+AM_CFLAGS = -Wall -Werror
+LDFLAGS = -static  # for speed and Valgrind
+LDADD = $(top_builddir)/src/libjansson.la

test/bin/json_process.c

@@ -0,0 +1,72 @@
+/*
+ * Copyright (c) 2009, 2010 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 <stdio.h>
+#include <stdlib.h>
+#include <jansson.h>
+
+static int getenv_int(const char *name)
+{
+    char *value, *end;
+    long result;
+
+    value = getenv(name);
+    if(!value)
+        return 0;
+
+    result = strtol(value, &end, 10);
+    if(*end != '\0')
+        return 0;
+
+    return (int)result;
+}
+
+int main(int argc, char *argv[])
+{
+    int indent = 0;
+    unsigned int flags = 0;
+
+    json_t *json;
+    json_error_t error;
+
+    if(argc != 1) {
+        fprintf(stderr, "usage: %s\n", argv[0]);
+        return 2;
+    }
+
+    indent = getenv_int("JSON_INDENT");
+    if(indent < 0 || indent > 255) {
+        fprintf(stderr, "invalid value for JSON_INDENT: %d\n", indent);
+        return 2;
+    }
+
+    if(indent > 0)
+        flags |= JSON_INDENT(indent);
+
+    if(getenv_int("JSON_COMPACT") > 0)
+        flags |= JSON_COMPACT;
+
+    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(!json) {
+        fprintf(stderr, "%d\n%s\n", error.line, error.text);
+        return 1;
+    }
+
+    json_dumpf(json, stdout, flags);
+    json_decref(json);
+
+    return 0;
+}

test/json-compare.py

@@ -1,45 +0,0 @@
-#!/usr/bin/python
-#
-# 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.
-
-import sys
-try:
-    import json
-except ImportError:
-    import simplejson as json
-
-def load(filename):
-    try:
-        jsonfile = open(filename)
-    except IOError, err:
-        print >>sys.stderr, "unable to load %s: %s" % \
-            (filename, err.strerror)
-        sys.exit(1)
-
-    try:
-        jsondata = json.load(jsonfile)
-    except ValueError, err:
-        print "%s is malformed: %s" % (filename, err)
-        sys.exit(1)
-    finally:
-        jsonfile.close()
-
-    return jsondata
-
-def main():
-    if len(sys.argv) != 3:
-        print >>sys.stderr, "usage: %s json1 json2" % sys.argv[0]
-        return 2
-
-    json1 = load(sys.argv[1])
-    json2 = load(sys.argv[2])
-    if json1 == json2:
-        return 0
-    else:
-        return 1
-
-if __name__ == '__main__':
-    sys.exit(main() or 0)

test/load_file_dump_file.c

@@ -1,31 +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.
- */
-
-#include <stdio.h>
-#include <jansson.h>
-
-int main(int argc, char *argv[])
-{
-    json_t *json;
-    json_error_t error;
-
-    if(argc != 3) {
-        fprintf(stderr, "usage: %s infile outfile\n", argv[0]);
-        return 2;
-    }
-
-    json = json_load_file(argv[1], &error);
-    if(!json) {
-        fprintf(stderr, "%d\n%s\n", error.line, error.text);
-        return 1;
-    }
-
-    json_dump_file(json, argv[2], 0);
-    json_decref(json);
-
-    return 0;
-}

test/loadf_dumpf.c

@@ -1,33 +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.
- */
-
-#include <stdio.h>
-#include <jansson.h>
-
-int main(int argc, char *argv[])
-{
-    json_t *json;
-    json_error_t error;
-
-    if(argc != 1) {
-        fprintf(stderr, "usage: %s\n", argv[0]);
-        return 2;
-    }
-
-    json = json_loadf(stdin, &error);
-    if(!json) {
-        fprintf(stderr, "%d\n%s\n", error.line, error.text);
-        return 1;
-    }
-
-    /* loadf_dumpf indents, others don't, so dumping with and without
-       indenting is tested */
-    json_dumpf(json, stdout, JSON_INDENT(4));
-    json_decref(json);
-
-    return 0;
-}

test/loads_dumps.c

@@ -1,47 +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.
- */
-
-#include <stdio.h>
-#include <stdlib.h>
-#include <jansson.h>
-
-#define BUFFER_SIZE (256 * 1024)
-
-int main(int argc, char *argv[])
-{
-    json_t *json;
-    json_error_t error;
-    int count;
-    char buffer[BUFFER_SIZE];
-    char *result;
-
-    if(argc != 1) {
-        fprintf(stderr, "usage: %s\n", argv[0]);
-        return 2;
-    }
-
-    count = fread(buffer, 1, BUFFER_SIZE, stdin);
-    if(count < 0 || count >= BUFFER_SIZE) {
-      fprintf(stderr, "unable to read input\n");
-      return 1;
-    }
-    buffer[count] = '\0';
-
-    json = json_loads(buffer, &error);
-    if(!json) {
-        fprintf(stderr, "%d\n%s\n", error.line, error.text);
-        return 1;
-    }
-
-    result = json_dumps(json, 0);
-    json_decref(json);
-
-    puts(result);
-    free(result);
-
-    return 0;
-}

test/run-suites

@@ -0,0 +1,46 @@
+#!/bin/sh
+
+while [ -n "$1" ]; do
+    suite=$1
+    if [ -x $top_srcdir/test/suites/$suite/run ]; then
+        SUITES="$SUITES $suite"
+    else
+        echo "No such suite: $suite"
+        exit 1
+    fi
+    shift
+done
+
+if [ -z "$SUITES" ]; then
+    suitedirs=$top_srcdir/test/suites/*
+    for suitedir in $suitedirs; do
+        if [ -d $suitedir ]; then
+            SUITES="$SUITES `basename $suitedir`"
+        fi
+    done
+fi
+
+export suites_srcdir=$top_srcdir/test/suites
+export suites_builddir=suites
+export scriptdir=$top_srcdir/test/scripts
+export logdir=logs
+export bindir=bin
+
+passed=0
+failed=0
+for suite in $SUITES; do
+    echo "Suite: $suite"
+    if $suites_srcdir/$suite/run $suite; then
+        passed=$(($passed+1))
+    else
+        failed=$(($failed+1))
+    fi
+done
+
+if [ $failed -gt 0 ]; then
+    echo "$failed of $((passed+failed)) test suites failed"
+    exit 1
+else
+    echo "$passed test suites passed"
+    rm -rf $logdir
+fi

test/run-test

@@ -1,54 +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.
-
-VALGRIND_CMDLINE="valgrind --leak-check=full --show-reachable=yes --track-origins=yes -q"
-
-run_testprog() {
-    local prog=$1
-    local prefix=$2
-    if [ -n "$VALGRIND" ]; then
-        local runner="$VALGRIND_CMDLINE "
-    fi
-
-    case "$prog" in
-        load_file_dump_file)
-            $runner./$prog \
-                $prefix.in \
-                $prefix.$prog.stdout \
-                2>$prefix.$prog.stderr
-            ;;
-        *)
-            $runner./$prog \
-                <$prefix.in \
-                >$prefix.$prog.stdout \
-                2>$prefix.$prog.stderr
-            ;;
-    esac
-
-    if [ -n "$VALGRIND" ]; then
-        # Check for Valgrind error output. The valgrind option
-        # --error-exitcode is not enough because Valgrind doesn't
-        # think unfreed allocs are errors.
-        if grep -E -q '^==[0-9]+== ' $prefix.$prog.stderr; then
-            echo "### $prefix ($prog) failed:" >&2
-            echo "valgrind detected an error" >&2
-            echo "for details, see test/$prefix.$prog.stderr" >&2
-            exit 1
-        fi
-    fi
-}
-
-for testfile in $TESTFILES; do
-    tmpdir="testlogs/`basename $testfile`"
-    rm -rf $tmpdir
-    mkdir -p $tmpdir
-    ${srcdir}/split-testfile.py $testfile $tmpdir | while read name; do
-        run_test loadf_dumpf $tmpdir/$name
-        run_test loads_dumps $tmpdir/$name
-        run_test load_file_dump_file $tmpdir/$name
-        echo -n '.'
-    done || exit 1
-    echo
-done

test/scripts/run-tests.sh

@@ -0,0 +1,71 @@
+# Copyright (c) 2009, 2010 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.
+
+json_process=$bindir/json_process
+
+suite_name=$1
+suite_srcdir=$suites_srcdir/$suite_name
+suite_builddir=$suites_builddir/$suite_name
+suite_log=$logdir/$suite_name
+
+
+[ -z "$VERBOSE" ] && VERBOSE=0
+
+. $scriptdir/valgrind.sh
+
+rm -rf $suite_log
+mkdir -p $suite_log
+
+for test_path in $suite_srcdir/*; do
+    test_name=$(basename $test_path)
+    test_builddir=$suite_builddir/$test_name
+    test_log=$suite_log/$test_name
+
+    [ "$test_name" = "run" ] && continue
+    is_test || continue
+
+    rm -rf $test_log
+    mkdir -p $test_log
+    if [ $VERBOSE -eq 1 ]; then
+        echo -n "$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
+done
+
+if [ $VERBOSE -eq 0 ]; then
+    echo
+fi
+
+if [ -n "$(ls -A $suite_log)" ]; then
+    for test_log in $suite_log/*; do
+        test_name=$(basename $test_log)
+        test_path=$suite_srcdir/$test_name
+        echo "================================================================="
+        echo "$suite_name/$test_name"
+        echo "================================================================="
+        show_error
+        echo
+    done
+    echo "================================================================="
+    exit 1
+else
+    rm -rf $suite_log
+fi

test/scripts/valgrind.sh

@@ -0,0 +1,35 @@
+# Copyright (c) 2009, 2010 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.
+
+[ -z "$VALGRIND" ] && VALGRIND=0
+
+VALGRIND_CMDLINE="valgrind --leak-check=full --show-reachable=yes --track-origins=yes -q"
+
+if [ $VALGRIND -eq 1 ]; then
+    test_runner="$VALGRIND_CMDLINE"
+    json_process="$VALGRIND_CMDLINE $json_process"
+else
+    test_runner=""
+fi
+
+valgrind_check() {
+    if [ $VALGRIND -eq 1 ]; then
+        # Check for Valgrind error output. The valgrind option
+        # --error-exitcode is not enough because Valgrind doesn't
+        # think unfreed allocs are errors.
+        if grep -E -q '^==[0-9]+== ' $1; then
+            touch $test_log/valgrind_error
+            return 1
+        fi
+    fi
+}
+
+valgrind_show_error() {
+    if [ $VALGRIND -eq 1 -a -f $test_log/valgrind_error ]; then
+        echo "valgrind detected an error"
+        return 0
+    fi
+    return 1
+}

test/split-testfile.py

@@ -1,54 +0,0 @@
-#!/usr/bin/python
-#
-# 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.
-
-import os
-import sys
-
-def open_files(outdir, i, name):
-    basename = '%02d_%s' % (i, name)
-    print basename
-    input_path = os.path.join(outdir, basename + '.in')
-    output_path = os.path.join(outdir, basename + '.out')
-    return open(input_path, 'w'), open(output_path, 'w')
-
-def main():
-    if len(sys.argv) != 3:
-        print 'usage: %s input-file output-directory' % sys.argv[0]
-        return 2
-
-    infile = os.path.normpath(sys.argv[1])
-    outdir = os.path.normpath(sys.argv[2])
-
-    if not os.path.exists(outdir):
-        print >>sys.stderr, 'output directory %r does not exist!' % outdir
-        return 1
-
-    n = 0
-    current = None
-    input, output = None, None
-
-    for line in open(infile):
-        if line.startswith('==== '):
-            n += 1
-            if input is not None and output is not None:
-                input.close()
-                output.close()
-            input, output = open_files(outdir, n, line[5:line.find(' ====\n')])
-            current = input
-        elif line == '====\n':
-            current = output
-        else:
-            current.write(line)
-
-    if input is not None and output is not None:
-        input.close()
-        output.close()
-
-    print >>sys.stderr, "%s: %d test cases" % (infile, n)
-
-if __name__ == '__main__':
-    sys.exit(main() or 0)

test/suites/Makefile.am

@@ -0,0 +1,2 @@
+SUBDIRS = api
+EXTRA_DIST = invalid invalid-strip invalid-unicode valid valid-strip

test/suites/api/Makefile.am

@@ -0,0 +1,24 @@
+EXTRA_DIST = run
+
+check_PROGRAMS = \
+	test_array \
+	test_equal \
+	test_copy \
+	test_dump \
+	test_load \
+	test_simple \
+	test_number \
+	test_object
+
+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_load_SOURCES = test_load.c util.h
+test_simple_SOURCES = test_simple.c util.h
+test_number_SOURCES = test_number.c util.h
+test_object_SOURCES = test_object.c util.h
+
+AM_CPPFLAGS = -I$(top_srcdir)/src
+AM_CFLAGS = -Wall -Werror
+LDFLAGS = -static  # for speed and Valgrind
+LDADD = $(top_builddir)/src/libjansson.la

test/suites/api/check-exports

@@ -0,0 +1,90 @@
+#!/bin/sh
+
+# This tests checks that the libjansson.so exports the correct
+# symbols.
+
+# The list of symbols that the shared object should export
+sort >$test_log/exports <<EOF
+json_delete
+json_true
+json_false
+json_null
+json_string
+json_string_nocheck
+json_string_value
+json_string_set
+json_string_set_nocheck
+json_integer
+json_integer_value
+json_integer_set
+json_real
+json_real_value
+json_real_set
+json_number_value
+json_array
+json_array_size
+json_array_get
+json_array_set_new
+json_array_append_new
+json_array_insert_new
+json_array_remove
+json_array_clear
+json_array_extend
+json_object
+json_object_size
+json_object_get
+json_object_set_new
+json_object_set_new_nocheck
+json_object_del
+json_object_clear
+json_object_update
+json_object_iter
+json_object_iter_at
+json_object_iter_next
+json_object_iter_key
+json_object_iter_value
+json_object_iter_set_new
+json_dumps
+json_dumpf
+json_dump_file
+json_loads
+json_loadf
+json_load_file
+json_equal
+json_copy
+json_deep_copy
+EOF
+
+# The list of functions are not exported in the library because they
+# are macros or static inline functions. This is only the make the
+# list complete, there are not used by the test.
+sort >$test_log/macros_or_inline <<EOF
+json_typeof
+json_incref
+json_decref
+json_is_object
+json_is_object
+json_is_array
+json_is_string
+json_is_integer
+json_is_real
+json_is_true
+json_is_false
+json_is_null
+json_is_number
+json_is_boolean
+json_array_set
+json_array_append
+json_array_insert
+json_object_set
+json_object_set_nocheck
+EOF
+
+SOFILE="../src/.libs/libjansson.so"
+
+nm -D $SOFILE | grep ' T ' | cut -d' ' -f3 | sort >$test_log/output
+
+if ! cmp -s $test_log/exports $test_log/output; then
+    diff -u $test_log/exports $test_log/output >&2
+    exit 1
+fi

test/suites/api/run

@@ -0,0 +1,36 @@
+#!/bin/sh
+#
+# Copyright (c) 2009, 2010 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() {
+    case "$test_name" in
+        *.c|check-exports)
+            return 0
+            ;;
+        *)
+            return 1
+            ;;
+    esac
+}
+
+run_test() {
+    if [ "$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 \
+            || return 1
+        valgrind_check $test_log/stderr || return 1
+    fi
+}
+
+show_error() {
+    valgrind_show_error && return
+    cat $test_log/stderr
+}
+
+. $top_srcdir/test/scripts/run-tests.sh

test/suites/api/test_array.c

@@ -0,0 +1,400 @@
+/*
+ * Copyright (c) 2009, 2010 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 "util.h"
+
+static void test_misc(void)
+{
+    json_t *array, *five, *seven, *value;
+    int i;
+
+    array = json_array();
+    five = json_integer(5);
+    seven = json_integer(7);
+
+    if(!array)
+        fail("unable to create array");
+    if(!five || !seven)
+        fail("unable to create integer");
+
+    if(json_array_size(array) != 0)
+        fail("empty array has nonzero size");
+
+    if(!json_array_append(array, NULL))
+        fail("able to append NULL");
+
+    if(json_array_append(array, five))
+        fail("unable to append");
+
+    if(json_array_size(array) != 1)
+        fail("wrong array size");
+
+    value = json_array_get(array, 0);
+    if(!value)
+        fail("unable to get item");
+    if(value != five)
+        fail("got wrong value");
+
+    if(json_array_append(array, seven))
+        fail("unable to append value");
+
+    if(json_array_size(array) != 2)
+        fail("wrong array size");
+
+    value = json_array_get(array, 1);
+    if(!value)
+        fail("unable to get item");
+    if(value != seven)
+        fail("got wrong value");
+
+    if(json_array_set(array, 0, seven))
+        fail("unable to set value");
+
+    if(!json_array_set(array, 0, NULL))
+        fail("able to set NULL");
+
+    if(json_array_size(array) != 2)
+        fail("wrong array size");
+
+    value = json_array_get(array, 0);
+    if(!value)
+        fail("unable to get item");
+    if(value != seven)
+        fail("got wrong value");
+
+    if(json_array_get(array, 2) != NULL)
+        fail("able to get value out of bounds");
+
+    if(!json_array_set(array, 2, seven))
+        fail("able to set value out of bounds");
+
+    for(i = 2; i < 30; i++) {
+        if(json_array_append(array, seven))
+            fail("unable to append value");
+
+        if(json_array_size(array) != i + 1)
+            fail("wrong array size");
+    }
+
+    for(i = 0; i < 30; i++) {
+        value = json_array_get(array, i);
+        if(!value)
+            fail("unable to get item");
+        if(value != seven)
+            fail("got wrong value");
+    }
+
+    if(json_array_set_new(array, 15, json_integer(123)))
+        fail("unable to set new value");
+
+    value = json_array_get(array, 15);
+    if(!json_is_integer(value) || json_integer_value(value) != 123)
+        fail("json_array_set_new works incorrectly");
+
+    if(!json_array_set_new(array, 15, NULL))
+        fail("able to set_new NULL value");
+
+    if(json_array_append_new(array, json_integer(321)))
+        fail("unable to append new value");
+
+    value = json_array_get(array, json_array_size(array) - 1);
+    if(!json_is_integer(value) || json_integer_value(value) != 321)
+        fail("json_array_append_new works incorrectly");
+
+    if(!json_array_append_new(array, NULL))
+        fail("able to append_new NULL value");
+
+    json_decref(five);
+    json_decref(seven);
+    json_decref(array);
+}
+
+static void test_insert(void)
+{
+    json_t *array, *five, *seven, *eleven, *value;
+    int i;
+
+    array = json_array();
+    five = json_integer(5);
+    seven = json_integer(7);
+    eleven = json_integer(11);
+
+    if(!array)
+        fail("unable to create array");
+    if(!five || !seven || !eleven)
+        fail("unable to create integer");
+
+
+    if(!json_array_insert(array, 1, five))
+        fail("able to insert value out of bounds");
+
+
+    if(json_array_insert(array, 0, five))
+        fail("unable to insert value in an empty array");
+
+    if(json_array_get(array, 0) != five)
+        fail("json_array_insert works incorrectly");
+
+    if(json_array_size(array) != 1)
+        fail("array size is invalid after insertion");
+
+
+    if(json_array_insert(array, 1, seven))
+        fail("unable to insert value at the end of an array");
+
+    if(json_array_get(array, 0) != five)
+        fail("json_array_insert works incorrectly");
+
+    if(json_array_get(array, 1) != seven)
+        fail("json_array_insert works incorrectly");
+
+    if(json_array_size(array) != 2)
+        fail("array size is invalid after insertion");
+
+
+    if(json_array_insert(array, 1, eleven))
+        fail("unable to insert value in the middle of an array");
+
+    if(json_array_get(array, 0) != five)
+        fail("json_array_insert works incorrectly");
+
+    if(json_array_get(array, 1) != eleven)
+        fail("json_array_insert works incorrectly");
+
+    if(json_array_get(array, 2) != seven)
+        fail("json_array_insert works incorrectly");
+
+    if(json_array_size(array) != 3)
+        fail("array size is invalid after insertion");
+
+
+    if(json_array_insert_new(array, 2, json_integer(123)))
+        fail("unable to insert value in the middle of an array");
+
+    value = json_array_get(array, 2);
+    if(!json_is_integer(value) || json_integer_value(value) != 123)
+        fail("json_array_insert_new works incorrectly");
+
+    if(json_array_size(array) != 4)
+        fail("array size is invalid after insertion");
+
+
+    for(i = 0; i < 20; i++) {
+        if(json_array_insert(array, 0, seven))
+            fail("unable to insert value at the begining of an array");
+    }
+
+    for(i = 0; i < 20; i++) {
+        if(json_array_get(array, i) != seven)
+            fail("json_aray_insert works incorrectly");
+    }
+
+    if(json_array_size(array) != 24)
+        fail("array size is invalid after loop insertion");
+
+    json_decref(five);
+    json_decref(seven);
+    json_decref(eleven);
+    json_decref(array);
+}
+
+static void test_remove(void)
+{
+    json_t *array, *five, *seven;
+
+    array = json_array();
+    five = json_integer(5);
+    seven = json_integer(7);
+
+    if(!array)
+        fail("unable to create array");
+    if(!five)
+        fail("unable to create integer");
+    if(!seven)
+        fail("unable to create integer");
+
+
+    if(!json_array_remove(array, 0))
+        fail("able to remove an unexisting index");
+
+
+    if(json_array_append(array, five))
+        fail("unable to append");
+
+    if(!json_array_remove(array, 1))
+        fail("able to remove an unexisting index");
+
+    if(json_array_remove(array, 0))
+        fail("unable to remove");
+
+    if(json_array_size(array) != 0)
+        fail("array size is invalid after removing");
+
+
+    if(json_array_append(array, five) ||
+       json_array_append(array, seven) ||
+       json_array_append(array, five) ||
+       json_array_append(array, seven))
+        fail("unable to append");
+
+    if(json_array_remove(array, 2))
+        fail("unable to remove");
+
+    if(json_array_size(array) != 3)
+        fail("array size is invalid after removing");
+
+    if(json_array_get(array, 0) != five ||
+       json_array_get(array, 1) != seven ||
+       json_array_get(array, 2) != seven)
+        fail("remove works incorrectly");
+
+    json_decref(five);
+    json_decref(seven);
+    json_decref(array);
+}
+
+static void test_clear(void)
+{
+    json_t *array, *five, *seven;
+    int i;
+
+    array = json_array();
+    five = json_integer(5);
+    seven = json_integer(7);
+
+    if(!array)
+        fail("unable to create array");
+    if(!five || !seven)
+        fail("unable to create integer");
+
+    for(i = 0; i < 10; i++) {
+        if(json_array_append(array, five))
+            fail("unable to append");
+    }
+    for(i = 0; i < 10; i++) {
+        if(json_array_append(array, seven))
+            fail("unable to append");
+    }
+
+    if(json_array_size(array) != 20)
+        fail("array size is invalid after appending");
+
+    if(json_array_clear(array))
+        fail("unable to clear");
+
+    if(json_array_size(array) != 0)
+        fail("array size is invalid after clearing");
+
+    json_decref(five);
+    json_decref(seven);
+    json_decref(array);
+}
+
+static void test_extend(void)
+{
+    json_t *array1, *array2, *five, *seven;
+    int i;
+
+    array1 = json_array();
+    array2 = json_array();
+    five = json_integer(5);
+    seven = json_integer(7);
+
+    if(!array1 || !array2)
+        fail("unable to create array");
+    if(!five || !seven)
+        fail("unable to create integer");
+
+    for(i = 0; i < 10; i++) {
+        if(json_array_append(array1, five))
+            fail("unable to append");
+    }
+    for(i = 0; i < 10; i++) {
+        if(json_array_append(array2, seven))
+            fail("unable to append");
+    }
+
+    if(json_array_size(array1) != 10 || json_array_size(array2) != 10)
+        fail("array size is invalid after appending");
+
+    if(json_array_extend(array1, array2))
+        fail("unable to extend");
+
+    for(i = 0; i < 10; i++) {
+        if(json_array_get(array1, i) != five)
+            fail("invalid array contents after extending");
+    }
+    for(i = 10; i < 20; i++) {
+        if(json_array_get(array1, i) != seven)
+            fail("invalid array contents after extending");
+    }
+
+    json_decref(five);
+    json_decref(seven);
+    json_decref(array1);
+    json_decref(array2);
+}
+
+static void test_circular()
+{
+    json_t *array1, *array2;
+
+    /* the simple cases are checked */
+
+    array1 = json_array();
+    if(!array1)
+        fail("unable to create array");
+
+    if(json_array_append(array1, array1) == 0)
+        fail("able to append self");
+
+    if(json_array_insert(array1, 0, array1) == 0)
+        fail("able to insert self");
+
+    if(json_array_append_new(array1, json_true()))
+        fail("failed to append true");
+
+    if(json_array_set(array1, 0, array1) == 0)
+        fail("able to set self");
+
+    json_decref(array1);
+
+
+    /* create circular references */
+
+    array1 = json_array();
+    array2 = json_array();
+    if(!array1 || !array2)
+        fail("unable to create array");
+
+    if(json_array_append(array1, array2) ||
+       json_array_append(array2, array1))
+        fail("unable to append");
+
+    /* circularity is detected when dumping */
+    if(json_dumps(array1, 0) != NULL)
+        fail("able to dump circulars");
+
+    /* decref twice to deal with the circular references */
+    json_decref(array1);
+    json_decref(array2);
+    json_decref(array1);
+}
+
+
+int main()
+{
+    test_misc();
+    test_insert();
+    test_remove();
+    test_clear();
+    test_extend();
+    test_circular();
+
+    return 0;
+}

test/suites/api/test_copy.c

@@ -0,0 +1,319 @@
+/*
+ * Copyright (c) 2009, 2010 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 <jansson.h>
+#include "util.h"
+
+static void test_copy_simple(void)
+{
+    json_t *value, *copy;
+
+    if(json_copy(NULL))
+        fail("copying NULL doesn't return NULL");
+
+    /* true */
+    value = json_true();
+    copy = json_copy(value);
+    if(value != copy)
+        fail("copying true failed");
+    json_decref(value);
+    json_decref(copy);
+
+    /* false */
+    value = json_false();
+    copy = json_copy(value);
+    if(value != copy)
+        fail("copying false failed");
+    json_decref(value);
+    json_decref(copy);
+
+    /* null */
+    value = json_null();
+    copy = json_copy(value);
+    if(value != copy)
+        fail("copying null failed");
+    json_decref(value);
+    json_decref(copy);
+
+    /* string */
+    value = json_string("foo");
+    if(!value)
+        fail("unable to create a string");
+    copy = json_copy(value);
+    if(!copy)
+        fail("unable to copy a string");
+    if(copy == value)
+        fail("copying a string doesn't copy");
+    if(!json_equal(copy, value))
+        fail("copying a string produces an inequal copy");
+    if(value->refcount != 1 || copy->refcount != 1)
+        fail("invalid refcounts");
+    json_decref(value);
+    json_decref(copy);
+
+    /* integer */
+    value = json_integer(543);
+    if(!value)
+        fail("unable to create an integer");
+    copy = json_copy(value);
+    if(!copy)
+        fail("unable to copy an integer");
+    if(copy == value)
+        fail("copying an integer doesn't copy");
+    if(!json_equal(copy, value))
+        fail("copying an integer produces an inequal copy");
+    if(value->refcount != 1 || copy->refcount != 1)
+        fail("invalid refcounts");
+    json_decref(value);
+    json_decref(copy);
+
+    /* real */
+    value = json_real(123e9);
+    if(!value)
+        fail("unable to create a real");
+    copy = json_copy(value);
+    if(!copy)
+        fail("unable to copy a real");
+    if(copy == value)
+        fail("copying a real doesn't copy");
+    if(!json_equal(copy, value))
+        fail("copying a real produces an inequal copy");
+    if(value->refcount != 1 || copy->refcount != 1)
+        fail("invalid refcounts");
+    json_decref(value);
+    json_decref(copy);
+}
+
+static void test_deep_copy_simple(void)
+{
+    json_t *value, *copy;
+
+    if(json_deep_copy(NULL))
+        fail("deep copying NULL doesn't return NULL");
+
+    /* true */
+    value = json_true();
+    copy = json_deep_copy(value);
+    if(value != copy)
+        fail("deep copying true failed");
+    json_decref(value);
+    json_decref(copy);
+
+    /* false */
+    value = json_false();
+    copy = json_deep_copy(value);
+    if(value != copy)
+        fail("deep copying false failed");
+    json_decref(value);
+    json_decref(copy);
+
+    /* null */
+    value = json_null();
+    copy = json_deep_copy(value);
+    if(value != copy)
+        fail("deep copying null failed");
+    json_decref(value);
+    json_decref(copy);
+
+    /* string */
+    value = json_string("foo");
+    if(!value)
+        fail("unable to create a string");
+    copy = json_deep_copy(value);
+    if(!copy)
+        fail("unable to deep copy a string");
+    if(copy == value)
+        fail("deep copying a string doesn't copy");
+    if(!json_equal(copy, value))
+        fail("deep copying a string produces an inequal copy");
+    if(value->refcount != 1 || copy->refcount != 1)
+        fail("invalid refcounts");
+    json_decref(value);
+    json_decref(copy);
+
+    /* integer */
+    value = json_integer(543);
+    if(!value)
+        fail("unable to create an integer");
+    copy = json_deep_copy(value);
+    if(!copy)
+        fail("unable to deep copy an integer");
+    if(copy == value)
+        fail("deep copying an integer doesn't copy");
+    if(!json_equal(copy, value))
+        fail("deep copying an integer produces an inequal copy");
+    if(value->refcount != 1 || copy->refcount != 1)
+        fail("invalid refcounts");
+    json_decref(value);
+    json_decref(copy);
+
+    /* real */
+    value = json_real(123e9);
+    if(!value)
+        fail("unable to create a real");
+    copy = json_deep_copy(value);
+    if(!copy)
+        fail("unable to deep copy a real");
+    if(copy == value)
+        fail("deep copying a real doesn't copy");
+    if(!json_equal(copy, value))
+        fail("deep copying a real produces an inequal copy");
+    if(value->refcount != 1 || copy->refcount != 1)
+        fail("invalid refcounts");
+    json_decref(value);
+    json_decref(copy);
+}
+
+static void test_copy_array(void)
+{
+    const char *json_array_text = "[1, \"foo\", 3.141592, {\"foo\": \"bar\"}]";
+
+    json_t *array, *copy;
+    unsigned int i;
+
+    array = json_loads(json_array_text, NULL);
+    if(!array)
+        fail("unable to parse an array");
+
+    copy = json_copy(array);
+    if(!copy)
+        fail("unable to copy an array");
+    if(copy == array)
+        fail("copying an array doesn't copy");
+    if(!json_equal(copy, array))
+        fail("copying an array produces an inequal copy");
+
+    for(i = 0; i < json_array_size(copy); i++)
+    {
+        if(json_array_get(array, i) != json_array_get(copy, i))
+            fail("copying an array modifies its elements");
+    }
+
+    json_decref(array);
+    json_decref(copy);
+}
+
+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;
+
+    array = json_loads(json_array_text, NULL);
+    if(!array)
+        fail("unable to parse an array");
+
+    copy = json_deep_copy(array);
+    if(!copy)
+        fail("unable to deep copy an array");
+    if(copy == array)
+        fail("deep copying an array doesn't copy");
+    if(!json_equal(copy, array))
+        fail("deep copying an array produces an inequal copy");
+
+    for(i = 0; i < json_array_size(copy); i++)
+    {
+        if(json_array_get(array, i) == json_array_get(copy, i))
+            fail("deep copying an array doesn't copy its elements");
+    }
+
+    json_decref(array);
+    json_decref(copy);
+}
+
+static void test_copy_object(void)
+{
+    const char *json_object_text =
+        "{\"foo\": \"bar\", \"a\": 1, \"b\": 3.141592, \"c\": [1,2,3,4]}";
+
+    json_t *object, *copy;
+    void *iter;
+
+    object = json_loads(json_object_text, NULL);
+    if(!object)
+        fail("unable to parse an object");
+
+    copy = json_copy(object);
+    if(!copy)
+        fail("unable to copy an object");
+    if(copy == object)
+        fail("copying an object doesn't copy");
+    if(!json_equal(copy, object))
+        fail("copying an object produces an inequal copy");
+
+    iter = json_object_iter(object);
+    while(iter)
+    {
+        const char *key;
+        json_t *value1, *value2;
+
+        key = json_object_iter_key(iter);
+        value1 = json_object_iter_value(iter);
+        value2 = json_object_get(copy, key);
+
+        if(value1 != value2)
+            fail("deep copying an object modifies its items");
+
+        iter = json_object_iter_next(object, iter);
+    }
+
+    json_decref(object);
+    json_decref(copy);
+}
+
+static void test_deep_copy_object(void)
+{
+    const char *json_object_text =
+        "{\"foo\": \"bar\", \"a\": 1, \"b\": 3.141592, \"c\": [1,2,3,4]}";
+
+    json_t *object, *copy;
+    void *iter;
+
+    object = json_loads(json_object_text, NULL);
+    if(!object)
+        fail("unable to parse an object");
+
+    copy = json_deep_copy(object);
+    if(!copy)
+        fail("unable to deep copy an object");
+    if(copy == object)
+        fail("deep copying an object doesn't copy");
+    if(!json_equal(copy, object))
+        fail("deep copying an object produces an inequal copy");
+
+    iter = json_object_iter(object);
+    while(iter)
+    {
+        const char *key;
+        json_t *value1, *value2;
+
+        key = json_object_iter_key(iter);
+        value1 = json_object_iter_value(iter);
+        value2 = json_object_get(copy, key);
+
+        if(value1 == value2)
+            fail("deep copying an object doesn't copy its items");
+
+        iter = json_object_iter_next(object, iter);
+    }
+
+    json_decref(object);
+    json_decref(copy);
+}
+
+int main()
+{
+    test_copy_simple();
+    test_deep_copy_simple();
+    test_copy_array();
+    test_deep_copy_array();
+    test_copy_object();
+    test_deep_copy_object();
+    return 0;
+}

test/suites/api/test_dump.c

@@ -0,0 +1,91 @@
+/*
+ * Copyright (c) 2009, 2010 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"
+
+int main()
+{
+    json_t *json;
+    char *result;
+
+    /* Encode an empty object/array, add an item, encode again */
+
+    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);
+
+    /* 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 = 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);
+
+    return 0;
+}

test/suites/api/test_equal.c

@@ -0,0 +1,190 @@
+/*
+ * Copyright (c) 2009, 2010 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 "util.h"
+
+static void test_equal_simple()
+{
+    json_t *value1, *value2;
+
+    if(json_equal(NULL, NULL))
+        fail("json_equal fails for two NULLs");
+
+    value1 = json_true();
+    if(json_equal(value1, NULL) || json_equal(NULL, value1))
+        fail("json_equal fails for NULL");
+
+    /* this covers true, false and null as they are singletons */
+    if(!json_equal(value1, value1))
+        fail("identical objects are not equal");
+    json_decref(value1);
+
+    /* integer */
+    value1 = json_integer(1);
+    value2 = json_integer(1);
+    if(!value1 || !value2)
+        fail("unable to create integers");
+    if(!json_equal(value1, value2))
+        fail("json_equal fails for two equal integers");
+    json_decref(value2);
+
+    value2 = json_integer(2);
+    if(!value2)
+        fail("unable to create an integer");
+    if(json_equal(value1, value2))
+        fail("json_equal fails for two inequal integers");
+
+    json_decref(value1);
+    json_decref(value2);
+
+    /* real */
+    value1 = json_real(1.2);
+    value2 = json_real(1.2);
+    if(!value1 || !value2)
+        fail("unable to create reals");
+    if(!json_equal(value1, value2))
+        fail("json_equal fails for two equal reals");
+    json_decref(value2);
+
+    value2 = json_real(3.141592);
+    if(!value2)
+        fail("unable to create an real");
+    if(json_equal(value1, value2))
+        fail("json_equal fails for two inequal reals");
+
+    json_decref(value1);
+    json_decref(value2);
+
+    /* string */
+    value1 = json_string("foo");
+    value2 = json_string("foo");
+    if(!value1 || !value2)
+        fail("unable to create strings");
+    if(!json_equal(value1, value2))
+        fail("json_equal fails for two equal strings");
+    json_decref(value2);
+
+    value2 = json_string("bar");
+    if(!value2)
+        fail("unable to create an string");
+    if(json_equal(value1, value2))
+        fail("json_equal fails for two inequal strings");
+
+    json_decref(value1);
+    json_decref(value2);
+}
+
+static void test_equal_array()
+{
+    json_t *array1, *array2;
+
+    array1 = json_array();
+    array2 = json_array();
+    if(!array1 || !array2)
+        fail("unable to create arrays");
+
+    if(!json_equal(array1, array2))
+        fail("json_equal fails for two empty arrays");
+
+    json_array_append_new(array1, json_integer(1));
+    json_array_append_new(array2, json_integer(1));
+    json_array_append_new(array1, json_string("foo"));
+    json_array_append_new(array2, json_string("foo"));
+    json_array_append_new(array1, json_integer(2));
+    json_array_append_new(array2, json_integer(2));
+    if(!json_equal(array1, array2))
+        fail("json_equal fails for two equal arrays");
+
+    json_array_remove(array2, 2);
+    if(json_equal(array1, array2))
+        fail("json_equal fails for two inequal arrays");
+
+    json_array_append_new(array2, json_integer(3));
+    if(json_equal(array1, array2))
+        fail("json_equal fails for two inequal arrays");
+
+    json_decref(array1);
+    json_decref(array2);
+}
+
+static void test_equal_object()
+{
+    json_t *object1, *object2;
+
+    object1 = json_object();
+    object2 = json_object();
+    if(!object1 || !object2)
+        fail("unable to create objects");
+
+    if(!json_equal(object1, object2))
+        fail("json_equal fails for two empty objects");
+
+    json_object_set_new(object1, "a", json_integer(1));
+    json_object_set_new(object2, "a", json_integer(1));
+    json_object_set_new(object1, "b", json_string("foo"));
+    json_object_set_new(object2, "b", json_string("foo"));
+    json_object_set_new(object1, "c", json_integer(2));
+    json_object_set_new(object2, "c", json_integer(2));
+    if(!json_equal(object1, object2))
+        fail("json_equal fails for two equal objects");
+
+    json_object_del(object2, "c");
+    if(json_equal(object1, object2))
+        fail("json_equal fails for two inequal objects");
+
+    json_object_set_new(object2, "c", json_integer(3));
+    if(json_equal(object1, object2))
+        fail("json_equal fails for two inequal objects");
+
+    json_object_del(object2, "c");
+    json_object_set_new(object2, "d", json_integer(2));
+    if(json_equal(object1, object2))
+        fail("json_equal fails for two inequal objects");
+
+    json_decref(object1);
+    json_decref(object2);
+}
+
+static void test_equal_complex()
+{
+    json_t *value1, *value2;
+
+    const char *complex_json =
+"{"
+"    \"integer\": 1, "
+"    \"real\": 3.141592, "
+"    \"string\": \"foobar\", "
+"    \"true\": true, "
+"    \"object\": {"
+"        \"array-in-object\": [1,true,\"foo\",{}],"
+"        \"object-in-object\": {\"foo\": \"bar\"}"
+"    },"
+"    \"array\": [\"foo\", false, null, 1.234]"
+"}";
+
+    value1 = json_loads(complex_json, NULL);
+    value2 = json_loads(complex_json, NULL);
+    if(!value1 || !value2)
+        fail("unable to parse JSON");
+    if(!json_equal(value1, value2))
+        fail("json_equal fails for two inequal strings");
+
+    json_decref(value1);
+    json_decref(value2);
+
+    /* TODO: There's no negative test case here */
+}
+
+int main()
+{
+    test_equal_simple();
+    test_equal_array();
+    test_equal_object();
+    test_equal_complex();
+    return 0;
+}

test/suites/api/test_load.c

@@ -0,0 +1,24 @@
+/*
+ * Copyright (c) 2009, 2010 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"
+
+int main()
+{
+    json_t *json;
+    json_error_t error;
+
+    json = json_load_file("/path/to/nonexistent/file.json", &error);
+    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;
+}

test/suites/api/test_number.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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/api/test_object.c

@@ -0,0 +1,451 @@
+/*
+ * Copyright (c) 2009, 2010 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 test_clear()
+{
+    json_t *object, *ten;
+
+    object = json_object();
+    ten = json_integer(10);
+
+    if(!object)
+        fail("unable to create object");
+    if(!ten)
+        fail("unable to create integer");
+
+    if(json_object_set(object, "a", ten) ||
+       json_object_set(object, "b", ten) ||
+       json_object_set(object, "c", ten) ||
+       json_object_set(object, "d", ten) ||
+       json_object_set(object, "e", ten))
+        fail("unable to set value");
+
+    if(json_object_size(object) != 5)
+        fail("invalid size");
+
+    json_object_clear(object);
+
+    if(json_object_size(object) != 0)
+        fail("invalid size after clear");
+
+    json_decref(ten);
+    json_decref(object);
+}
+
+static void test_update()
+{
+    json_t *object, *other, *nine, *ten;
+
+    object = json_object();
+    other = json_object();
+
+    nine = json_integer(9);
+    ten = json_integer(10);
+
+    if(!object || !other)
+        fail("unable to create object");
+    if(!nine || !ten)
+        fail("unable to create integer");
+
+
+    /* update an empty object with an empty object */
+
+    if(json_object_update(object, other))
+        fail("unable to update an emtpy object with an empty object");
+
+    if(json_object_size(object) != 0)
+        fail("invalid size after update");
+
+    if(json_object_size(other) != 0)
+        fail("invalid size for updater after update");
+
+
+    /* update an empty object with a nonempty object */
+
+    if(json_object_set(other, "a", ten) ||
+       json_object_set(other, "b", ten) ||
+       json_object_set(other, "c", ten) ||
+       json_object_set(other, "d", ten) ||
+       json_object_set(other, "e", ten))
+        fail("unable to set value");
+
+    if(json_object_update(object, other))
+        fail("unable to update an empty object");
+
+    if(json_object_size(object) != 5)
+        fail("invalid size after update");
+
+    if(json_object_get(object, "a") != ten ||
+       json_object_get(object, "b") != ten ||
+       json_object_get(object, "c") != ten ||
+       json_object_get(object, "d") != ten ||
+       json_object_get(object, "e") != ten)
+        fail("update works incorrectly");
+
+
+    /* perform the same update again */
+
+    if(json_object_update(object, other))
+        fail("unable to update an empty object");
+
+    if(json_object_size(object) != 5)
+        fail("invalid size after update");
+
+    if(json_object_get(object, "a") != ten ||
+       json_object_get(object, "b") != ten ||
+       json_object_get(object, "c") != ten ||
+       json_object_get(object, "d") != ten ||
+       json_object_get(object, "e") != ten)
+        fail("update works incorrectly");
+
+
+    /* update a nonempty object with a nonempty object with both old
+       and new keys */
+
+    if(json_object_clear(other))
+        fail("clear failed");
+
+    if(json_object_set(other, "a", nine) ||
+       json_object_set(other, "b", nine) ||
+       json_object_set(other, "f", nine) ||
+       json_object_set(other, "g", nine) ||
+       json_object_set(other, "h", nine))
+        fail("unable to set value");
+
+    if(json_object_update(object, other))
+        fail("unable to update a nonempty object");
+
+    if(json_object_size(object) != 8)
+        fail("invalid size after update");
+
+    if(json_object_get(object, "a") != nine ||
+       json_object_get(object, "b") != nine ||
+       json_object_get(object, "f") != nine ||
+       json_object_get(object, "g") != nine ||
+       json_object_get(object, "h") != nine)
+        fail("update works incorrectly");
+
+    json_decref(nine);
+    json_decref(ten);
+    json_decref(other);
+    json_decref(object);
+}
+
+static void test_circular()
+{
+    json_t *object1, *object2;
+
+    object1 = json_object();
+    object2 = json_object();
+    if(!object1 || !object2)
+        fail("unable to create object");
+
+    /* the simple case is checked */
+    if(json_object_set(object1, "a", object1) == 0)
+        fail("able to set self");
+
+    /* create circular references */
+    if(json_object_set(object1, "a", object2) ||
+       json_object_set(object2, "a", object1))
+        fail("unable to set value");
+
+    /* circularity is detected when dumping */
+    if(json_dumps(object1, 0) != NULL)
+        fail("able to dump circulars");
+
+    /* decref twice to deal with the circular references */
+    json_decref(object1);
+    json_decref(object2);
+    json_decref(object1);
+}
+
+static void test_set_nocheck()
+{
+    json_t *object, *string;
+
+    object = json_object();
+    string = json_string("bar");
+
+    if(!object)
+        fail("unable to create object");
+    if(!string)
+        fail("unable to create string");
+
+    if(json_object_set_nocheck(object, "foo", string))
+        fail("json_object_set_nocheck failed");
+    if(json_object_get(object, "foo") != string)
+        fail("json_object_get after json_object_set_nocheck failed");
+
+    /* invalid UTF-8 in key */
+    if(json_object_set_nocheck(object, "a\xefz", string))
+        fail("json_object_set_nocheck failed for invalid UTF-8");
+    if(json_object_get(object, "a\xefz") != string)
+        fail("json_object_get after json_object_set_nocheck failed");
+
+    if(json_object_set_new_nocheck(object, "bax", json_integer(123)))
+        fail("json_object_set_new_nocheck failed");
+    if(json_integer_value(json_object_get(object, "bax")) != 123)
+        fail("json_object_get after json_object_set_new_nocheck failed");
+
+    /* invalid UTF-8 in key */
+    if(json_object_set_new_nocheck(object, "asdf\xfe", json_integer(321)))
+        fail("json_object_set_new_nocheck failed for invalid UTF-8");
+    if(json_integer_value(json_object_get(object, "asdf\xfe")) != 321)
+        fail("json_object_get after json_object_set_new_nocheck failed");
+
+    json_decref(string);
+    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;
+
+    object = json_object();
+    string = json_string("test");
+    other_string = json_string("other");
+
+    if(!object)
+        fail("unable to create object");
+    if(!string || !other_string)
+        fail("unable to create string");
+
+    if(json_object_get(object, "a"))
+        fail("value for nonexisting key");
+
+    if(json_object_set(object, "a", string))
+        fail("unable to set value");
+
+    if(!json_object_set(object, NULL, string))
+        fail("able to set NULL key");
+
+    if(!json_object_set(object, "a", NULL))
+        fail("able to set NULL value");
+
+    /* invalid UTF-8 in key */
+    if(!json_object_set(object, "a\xefz", string))
+        fail("able to set invalid unicode key");
+
+    value = json_object_get(object, "a");
+    if(!value)
+        fail("no value for existing key");
+    if(value != string)
+        fail("got different value than what was added");
+
+    /* "a", "lp" and "px" collide in a five-bucket hashtable */
+    if(json_object_set(object, "b", string) ||
+       json_object_set(object, "lp", string) ||
+       json_object_set(object, "px", string))
+        fail("unable to set value");
+
+    value = json_object_get(object, "a");
+    if(!value)
+        fail("no value for existing key");
+    if(value != string)
+        fail("got different value than what was added");
+
+    if(json_object_set(object, "a", other_string))
+        fail("unable to replace an existing key");
+
+    value = json_object_get(object, "a");
+    if(!value)
+        fail("no value for existing key");
+    if(value != other_string)
+        fail("got different value than what was set");
+
+    if(!json_object_del(object, "nonexisting"))
+        fail("able to delete a nonexisting key");
+
+    if(json_object_del(object, "px"))
+        fail("unable to delete an existing key");
+
+    if(json_object_del(object, "a"))
+        fail("unable to delete an existing key");
+
+    if(json_object_del(object, "lp"))
+        fail("unable to delete an existing key");
+
+
+    /* add many keys to initiate rehashing */
+
+    if(json_object_set(object, "a", string))
+        fail("unable to set value");
+
+    if(json_object_set(object, "lp", string))
+        fail("unable to set value");
+
+    if(json_object_set(object, "px", string))
+        fail("unable to set value");
+
+    if(json_object_set(object, "c", string))
+        fail("unable to set value");
+
+    if(json_object_set(object, "d", string))
+        fail("unable to set value");
+
+    if(json_object_set(object, "e", string))
+        fail("unable to set value");
+
+
+    if(json_object_set_new(object, "foo", json_integer(123)))
+        fail("unable to set new value");
+
+    value = json_object_get(object, "foo");
+    if(!json_is_integer(value) || json_integer_value(value) != 123)
+        fail("json_object_set_new works incorrectly");
+
+    if(!json_object_set_new(object, NULL, json_integer(432)))
+        fail("able to set_new NULL key");
+
+    if(!json_object_set_new(object, "foo", NULL))
+        fail("able to set_new NULL value");
+
+    json_decref(string);
+    json_decref(other_string);
+    json_decref(object);
+}
+
+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);
+}
+
+int main()
+{
+    test_misc();
+    test_clear();
+    test_update();
+    test_circular();
+    test_set_nocheck();
+    test_iterators();
+    test_preserve_order();
+
+    return 0;
+}

test/suites/api/test_simple.c

@@ -0,0 +1,185 @@
+/*
+ * Copyright (c) 2009, 2010 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 <jansson.h>
+#include "util.h"
+
+/* Call the simple functions not covered by other tests of the public API */
+int main()
+{
+    json_t *value;
+
+    value = json_integer(1);
+    if(json_typeof(value) != JSON_INTEGER)
+        fail("json_typeof failed");
+
+    if(json_is_object(value))
+        fail("json_is_object failed");
+
+    if(json_is_array(value))
+        fail("json_is_array failed");
+
+    if(json_is_string(value))
+        fail("json_is_string failed");
+
+    if(!json_is_integer(value))
+        fail("json_is_integer failed");
+
+    if(json_is_real(value))
+        fail("json_is_real failed");
+
+    if(!json_is_number(value))
+        fail("json_is_number failed");
+
+    if(json_is_true(value))
+        fail("json_is_true failed");
+
+    if(json_is_false(value))
+        fail("json_is_false failed");
+
+    if(json_is_boolean(value))
+        fail("json_is_boolean failed");
+
+    if(json_is_null(value))
+        fail("json_is_null failed");
+
+    json_decref(value);
+
+
+    value = json_string("foo");
+    if(!value)
+        fail("json_string failed");
+    if(strcmp(json_string_value(value), "foo"))
+        fail("invalid string value");
+
+    if(json_string_set(value, "bar"))
+        fail("json_string_set failed");
+    if(strcmp(json_string_value(value), "bar"))
+        fail("invalid string value");
+
+    json_decref(value);
+
+    value = json_string(NULL);
+    if(value)
+        fail("json_string(NULL) failed");
+
+    /* invalid UTF-8  */
+    value = json_string("a\xefz");
+    if(value)
+        fail("json_string(<invalid utf-8>) failed");
+
+    value = json_string_nocheck("foo");
+    if(!value)
+        fail("json_string_nocheck failed");
+    if(strcmp(json_string_value(value), "foo"))
+        fail("invalid string value");
+
+    if(json_string_set_nocheck(value, "bar"))
+        fail("json_string_set_nocheck failed");
+    if(strcmp(json_string_value(value), "bar"))
+        fail("invalid string value");
+
+    json_decref(value);
+
+    /* invalid UTF-8 */
+    value = json_string_nocheck("qu\xff");
+    if(!value)
+        fail("json_string_nocheck failed");
+    if(strcmp(json_string_value(value), "qu\xff"))
+        fail("invalid string value");
+
+    if(json_string_set_nocheck(value, "\xfd\xfe\xff"))
+        fail("json_string_set_nocheck failed");
+    if(strcmp(json_string_value(value), "\xfd\xfe\xff"))
+        fail("invalid string value");
+
+    json_decref(value);
+
+
+    value = json_integer(123);
+    if(!value)
+        fail("json_integer failed");
+    if(json_integer_value(value) != 123)
+        fail("invalid integer value");
+    if(json_number_value(value) != 123.0)
+        fail("invalid number value");
+
+    if(json_integer_set(value, 321))
+        fail("json_integer_set failed");
+    if(json_integer_value(value) != 321)
+        fail("invalid integer value");
+    if(json_number_value(value) != 321.0)
+        fail("invalid number value");
+
+    json_decref(value);
+
+    value = json_real(123.123);
+    if(!value)
+        fail("json_real failed");
+    if(json_real_value(value) != 123.123)
+        fail("invalid integer value");
+    if(json_number_value(value) != 123.123)
+        fail("invalid number value");
+
+    if(json_real_set(value, 321.321))
+        fail("json_real_set failed");
+    if(json_real_value(value) != 321.321)
+        fail("invalid real value");
+    if(json_number_value(value) != 321.321)
+        fail("invalid number value");
+
+    json_decref(value);
+
+    value = json_true();
+    if(!value)
+        fail("json_true failed");
+    json_decref(value);
+
+    value = json_false();
+    if(!value)
+        fail("json_false failed");
+    json_decref(value);
+
+    value = json_null();
+    if(!value)
+        fail("json_null failed");
+    json_decref(value);
+
+    /* Test reference counting on singletons (true, false, null) */
+    value = json_true();
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting true works incorrectly");
+    json_decref(value);
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting true works incorrectly");
+    json_incref(value);
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting true works incorrectly");
+
+    value = json_false();
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting false works incorrectly");
+    json_decref(value);
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting false works incorrectly");
+    json_incref(value);
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting false works incorrectly");
+
+    value = json_null();
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting null works incorrectly");
+    json_decref(value);
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting null works incorrectly");
+    json_incref(value);
+    if(value->refcount != (unsigned int)-1)
+      fail("refcounting null works incorrectly");
+
+    return 0;
+}

test/suites/api/util.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2009 Petri Lehtinen <petri@digip.org>
+ * Copyright (c) 2009, 2010 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.
@@ -8,10 +8,13 @@
 #ifndef TESTPROGS_UTIL_H
 #define TESTPROGS_UTIL_H
 
+#include <stdlib.h>
+
 #define fail(msg)                                                \
     do {                                                         \
-        fprintf(stderr, "%s:%d: %s\n", __FILE__, __LINE__, msg); \
-        return 1;                                                \
+        fprintf(stderr, "%s:%s:%d: %s\n",                        \
+                __FILE__, __FUNCTION__, __LINE__, msg);          \
+        exit(1);                                                 \
     } while(0)
 
 #endif

test/suites/encoding-flags/array/input

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

test/suites/encoding-flags/array/output

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

test/suites/encoding-flags/compact-array/env

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

test/suites/encoding-flags/compact-array/input

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

test/suites/encoding-flags/compact-array/output

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

test/suites/encoding-flags/compact-object/env

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

test/suites/encoding-flags/compact-object/input

@@ -0,0 +1 @@
+{"a": 1, "b": 2}

test/suites/encoding-flags/compact-object/output

@@ -0,0 +1 @@
+{"a":1,"b":2}

test/suites/encoding-flags/ensure-ascii/env

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

test/suites/encoding-flags/ensure-ascii/input

@@ -0,0 +1,8 @@
+[
+    "foo",
+    "å ä ö",
+    "foo åä",
+    "åä foo",
+    "å foo ä",
+    "clef g: 𝄞"
+]

test/suites/encoding-flags/ensure-ascii/output

@@ -0,0 +1 @@
+["foo", "\u00e5 \u00e4 \u00f6", "foo \u00e5\u00e4", "\u00e5\u00e4 foo", "\u00e5 foo \u00e4", "clef g: \ud834\udd1e"]

test/suites/encoding-flags/indent-array/env

@@ -0,0 +1 @@
+export JSON_INDENT=4

test/suites/encoding-flags/indent-array/input

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

test/suites/encoding-flags/indent-array/output

@@ -0,0 +1,4 @@
+[
+    1,
+    2
+]

test/suites/encoding-flags/indent-compact-array/env

@@ -0,0 +1,2 @@
+export JSON_INDENT=4
+export JSON_COMPACT=1

test/suites/encoding-flags/indent-compact-array/input

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

test/suites/encoding-flags/indent-compact-array/output

@@ -0,0 +1,4 @@
+[
+    1,
+    2
+]

test/suites/encoding-flags/indent-compact-object/env

@@ -0,0 +1,2 @@
+export JSON_INDENT=4
+export JSON_COMPACT=1

test/suites/encoding-flags/indent-compact-object/input

@@ -0,0 +1 @@
+{"a": 1, "b": 2}

test/suites/encoding-flags/indent-compact-object/output

@@ -0,0 +1,4 @@
+{
+    "a":1,
+    "b":2
+}

test/suites/encoding-flags/indent-object/env

@@ -0,0 +1 @@
+export JSON_INDENT=4

test/suites/encoding-flags/indent-object/input

@@ -0,0 +1 @@
+{"a": 1, "b": 2}

test/suites/encoding-flags/indent-object/output

@@ -0,0 +1,4 @@
+{
+    "a": 1,
+    "b": 2
+}

test/suites/encoding-flags/object/input

@@ -0,0 +1 @@
+{"a": 1, "b": 2}

test/suites/encoding-flags/object/output

@@ -0,0 +1 @@
+{"a": 1, "b": 2}

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

@@ -0,0 +1,32 @@
+#!/bin/sh
+#
+# Copyright (c) 2009, 2010 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 -d $test_path
+}
+
+run_test() {
+    (
+        if [ -f $test_path/env ]; then
+            . $test_path/env
+        fi
+        $json_process <$test_path/input >$test_log/stdout 2>$test_log/stderr
+    )
+    valgrind_check $test_log/stderr || return 1
+    cmp -s $test_path/output $test_log/stdout
+}
+
+show_error() {
+    valgrind_show_error && return
+
+    echo "EXPECTED OUTPUT:"
+    nl -bn $test_path/output
+    echo "ACTUAL OUTPUT:"
+    nl -bn $test_log/stdout
+}
+
+. $top_srcdir/test/scripts/run-tests.sh

test/suites/encoding-flags/sort-keys/env

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

test/suites/encoding-flags/sort-keys/input

@@ -0,0 +1 @@
+{"foo": 1, "bar": 2, "baz": 3, "quux": 4}

test/suites/encoding-flags/sort-keys/output

@@ -0,0 +1 @@
+{"bar": 2, "baz": 3, "foo": 1, "quux": 4}

test/suites/invalid-strip/apostrophe/error

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

test/suites/invalid-strip/apostrophe/input

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

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

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

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

@@ -0,0 +1 @@
+aå

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

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

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

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

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

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

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

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

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

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