summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2014-10-01tests: Add XPath sorting tests for move/append_bufferArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1035 99668b35-9821-0410-8761-19e4c4f06640
2014-10-01Disable document_order optimization after move/append_buffer.Arseny Kapoulkine
Moving nodes results in node order being different from order of allocated names/values; since move is O(1) we can't mark the moved nodes in a subtree so we have to disable the optimization for the entire document. Similarly, if a node is composed of multiple buffers, comparing nodes in different buffers does not result in meaningful order. Since we value correctness over performance, mark the entire document in these cases to disable sorting optimization. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1034 99668b35-9821-0410-8761-19e4c4f06640
2014-10-01tests: Add tests for copyless copy and related potential bugsArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1033 99668b35-9821-0410-8761-19e4c4f06640
2014-10-01Implement copyless copyArseny Kapoulkine
Now copying nodes or attributes does not copy names/values if the source strings are in a document buffer. As a result, several nodes can now share the same string in document buffer - to support this we 'taint' both source and destination with a special 'shared' bit. Tainting disables offset_debug() and fast-path document order comparison; it also prevents strcpy_insitu from reusing the document buffer memory for the copied node. The downsides include slower XPath queries in some (rare) cases and slightly higher memory consumption in some (rare) cases. XPath queries can execute slower if a lot of old nodes were copied to new nodes *and* a query only touches old nodes (so it used to benefit a lot from fast comparison path) *and* a query produces unsorted node sets that need to be sorted later (both are relatively rare). Higher memory consumption is possible if a lot of nodes were copied and all nodes (both new and old) have their contents modified 'in place' -- previously we could modify the old node in place and the new node required one allocation on copy, and now both nodes have to have their data allocated during modification. This should also be rare. On the bright side, in a lot of cases copying of string data can be avoided - this makes the copy much faster and the document now occupies less memory. For example, some uses of append_buffer are now actually slower compared to building up a document by copying a template from the same document and modifying the copy slightly. In one of the internal benchmarks copying is now 4x faster (the difference can be more dramatic with more string contents and less markup). git-svn-id: https://pugixml.googlecode.com/svn/trunk@1032 99668b35-9821-0410-8761-19e4c4f06640
2014-10-01Add header bit for 'name or value is shared' flagArseny Kapoulkine
This is required to make it possible to use a pointer to one of the buffers with the document data in nodes but keep offset_debug and (more importantly) XPath document order comparison optimization working. The change increases memory page alignment to 64 bytes (so requires +32 bytes for every page allocation, which should not be a problem - even with non-default 4k pages this is <1% extra cost, with default 32k pages the overhead is 0.1%) git-svn-id: https://pugixml.googlecode.com/svn/trunk@1031 99668b35-9821-0410-8761-19e4c4f06640
2014-10-01tests: Fix allocator.cpp indentationArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1030 99668b35-9821-0410-8761-19e4c4f06640
2014-09-28Remove redundant reference indirection for internal functions.Arseny Kapoulkine
Since xml_node/attribute are pointer wrappers it's cheaper to pass them by value. This makes XPath evaluation 4% faster and node printing 2% faster. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1029 99668b35-9821-0410-8761-19e4c4f06640
2014-09-28tests: Add a test for stackless copyArseny Kapoulkine
This test has previously caused a stack overflow on x86/MSVC. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1028 99668b35-9821-0410-8761-19e4c4f06640
2014-09-28Implement non-recursive node copyingArseny Kapoulkine
This makes node copying 6% faster, prevents it from ever running out of stack space and makes the profiling results more actionable for profilers that can't merge information from recursive calls. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1027 99668b35-9821-0410-8761-19e4c4f06640
2014-09-27contrib: Fix foreach.hpp for Boost 1.56.0Arseny Kapoulkine
Include range/iterator.hpp to avoid compatibility issues in the future. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1026 99668b35-9821-0410-8761-19e4c4f06640
2014-09-25Internal refactoring: rename xml_buffered_writer::write overloadsArseny Kapoulkine
Renames write to write_string and write_buffer to make it easier to distinguish between them in profiling runs and commit messages... git-svn-id: https://pugixml.googlecode.com/svn/trunk@1025 99668b35-9821-0410-8761-19e4c4f06640
2014-09-25Optimize xml_buffered_writer::write(char_t...)Arseny Kapoulkine
Make it easier for the compiler to generate good code by loading bufsize into a local once and returning new offset from flush(). This results in 7% performance gain. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1024 99668b35-9821-0410-8761-19e4c4f06640
2014-09-24Optimize xml_buffered_writer::write(const char_t*)Arseny Kapoulkine
Instead of computing the length and doing memcpy we now copy the head of the string into the buffer (like strcpy) and then use memcpy for tail if necessary. This results in 10-15% speedup for writing typical documents with a mix of short and long strings. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1023 99668b35-9821-0410-8761-19e4c4f06640
2014-09-23tests: Add one more XPath optimization testArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1022 99668b35-9821-0410-8761-19e4c4f06640
2014-09-23XPath: Optimize //name queries when possibleArseny Kapoulkine
//name means /descendant-or-self::node()/child::name, but we frequently can replace it with /descendant::name. This means we do not have to build up a temporary node set with all descendants that can lead to 3x speedups. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1021 99668b35-9821-0410-8761-19e4c4f06640
2014-09-23tests: More XPath testsArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1020 99668b35-9821-0410-8761-19e4c4f06640
2014-09-22Optimize and refactor node output implementation a bit (+5% perf gain)Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1019 99668b35-9821-0410-8761-19e4c4f06640
2014-09-21Optimize text_output_indentArseny Kapoulkine
We now precompute indent length and have a fast path for lengths 0..4 that avoids calling memcpy in a tight loop. This makes node output 20-30% faster if indentation is enabled. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1018 99668b35-9821-0410-8761-19e4c4f06640
2014-09-21tests: Fix PUGIXML_WCHAR_MODE buildArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1017 99668b35-9821-0410-8761-19e4c4f06640
2014-09-21tests: Add test for custom indentation stringsArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1016 99668b35-9821-0410-8761-19e4c4f06640
2014-09-21tests: Add a test for stackless write.Arseny Kapoulkine
This test previously caused a stack overflow on x86/MSVC. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1015 99668b35-9821-0410-8761-19e4c4f06640
2014-09-21Implement non-recursive node outputArseny Kapoulkine
This makes node output 3% faster, prevents it from ever running out of stack space and makes the profiling results more actionable for profilers that can't merge information from recursive calls. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1014 99668b35-9821-0410-8761-19e4c4f06640
2014-09-21Fix Jamrules.jam for Windows Kits 8.1Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1013 99668b35-9821-0410-8761-19e4c4f06640
2014-09-15CMake tweaks:Arseny Kapoulkine
- Include GNUInstallDirs which sets up standard install locations including lib64 for mulilib systems. - Make BUILD_SHARED_LIBS an option instead of a variable which is better for use in either the cmake-gui or ccmake gui interfaces. - Setup a destination for WIN32 runtime DLL's which is also helpful for MinGW installs. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1012 99668b35-9821-0410-8761-19e4c4f06640
2014-09-15Fix VC 14 warningsArseny Kapoulkine
Fixes C4458: declaration of 'var' hides class member git-svn-id: https://pugixml.googlecode.com/svn/trunk@1011 99668b35-9821-0410-8761-19e4c4f06640
2014-09-12tests: Add git-svn reviving scriptArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1010 99668b35-9821-0410-8761-19e4c4f06640
2014-09-04Merged pull request #7 from opoplawski/master.Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1009 99668b35-9821-0410-8761-19e4c4f06640
2014-09-03scripts: Add EXPORT pugixml-configOrion Poplawski
2014-08-26Unroll performance-critical loops 4xArseny Kapoulkine
Use a special macro that unrolls the loop body and uses static branch prediction to improve code generation. This increases performance across all data sets from benchmark; clang x64 is 10%-40% faster, clang x86 is 5%-20% faster, msvc is 5%-10% faster. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1008 99668b35-9821-0410-8761-19e4c4f06640
2014-08-25Use PUGI__SCANWHILE for strconv utilitiesArseny Kapoulkine
This makes it easier to optimize strconv. For consistency move all definitions of parser-internal macros to one place. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1007 99668b35-9821-0410-8761-19e4c4f06640
2014-08-25Rename ENDSWITH to PUGI__ENDSWITHArseny Kapoulkine
Also add it to #undef list at the end to avoid conflicts git-svn-id: https://pugixml.googlecode.com/svn/trunk@1006 99668b35-9821-0410-8761-19e4c4f06640
2014-08-11Move attribute name setup after pointer setup to handle exceptions betterArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1005 99668b35-9821-0410-8761-19e4c4f06640
2014-08-10docs: Add documentation for moving functionsArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1004 99668b35-9821-0410-8761-19e4c4f06640
2014-08-10tests: Add tests for node movementArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1003 99668b35-9821-0410-8761-19e4c4f06640
2014-08-10Implement node moving functions.Arseny Kapoulkine
The operations itself are O(1) since they just rearrange pointers. However, the validation step is O(logN) due to a sanity check to prevent recursive trees. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1002 99668b35-9821-0410-8761-19e4c4f06640
2014-08-10Refactor low-level node manipulation routines into separate functions.Arseny Kapoulkine
This should make moving implementation easier. git-svn-id: https://pugixml.googlecode.com/svn/trunk@1001 99668b35-9821-0410-8761-19e4c4f06640
2014-07-28Add BUILD_DEFINES parameter to CMakeLists.txtArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@1000 99668b35-9821-0410-8761-19e4c4f06640
2014-06-01Improve XPath allocator performanceArseny Kapoulkine
When allocating new pages, make sure that the page has at least 1/4 of the base page size free. This makes sure that we can do small allocations after big allocations (i.e. huge node lists) without doing a heap alloc. This is important because XPath stack code always reclaims extra pages after evaluating sub-expressions, so allocating a small chunk of memory and then rolling the state back is a common case (filtering a node list using a predicate usually does this). A better solution involves smarter allocation rollback strategy, but the implemented solution is simple and practical. git-svn-id: https://pugixml.googlecode.com/svn/trunk@999 99668b35-9821-0410-8761-19e4c4f06640
2014-05-04Add xpath_node_set::iterator that is the same as const_iterator.Arseny Kapoulkine
Exposing true mutable iterators allows the user to violate sorting order contract. However, some generic algorithms (i.e. Boost ForEach) require iterator methods to be present. git-svn-id: https://pugixml.googlecode.com/svn/trunk@998 99668b35-9821-0410-8761-19e4c4f06640
2014-04-02tests: Fix Mac OS X compilation.Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@997 99668b35-9821-0410-8761-19e4c4f06640
2014-02-28docs: Update release date to... todayv1.4Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@994 99668b35-9821-0410-8761-19e4c4f06640
2014-02-28docs: Regenerate HTML documentationArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@993 99668b35-9821-0410-8761-19e4c4f06640
2014-02-28docs: Fix samples compilationArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@992 99668b35-9821-0410-8761-19e4c4f06640
2014-02-26tests: Fix WinCE compilationArseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@991 99668b35-9821-0410-8761-19e4c4f06640
2014-02-25Revert automatic support for header-only mode since it creates problems with ↵Arseny Kapoulkine
qmake. Qmake treats all files that are #include-d as header files, even if the #include is guarded by an #ifdef. It looks like the only solution that allows for transparent header-only support based on preprocessor define involves moving the actual source into a separate header file and including this file in pugixml.cpp. Let's not do it yet. git-svn-id: https://pugixml.googlecode.com/svn/trunk@990 99668b35-9821-0410-8761-19e4c4f06640
2014-02-25docs: Add parse_trim_pcdata documentation, change release date to March 1st.Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@989 99668b35-9821-0410-8761-19e4c4f06640
2014-02-25tests: Add tests for parse_trim_pcdata.Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@988 99668b35-9821-0410-8761-19e4c4f06640
2014-02-25Add parse_trim_pcdata parse option.Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@987 99668b35-9821-0410-8761-19e4c4f06640
2014-02-23Fix clang build.Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@986 99668b35-9821-0410-8761-19e4c4f06640
2014-02-23Fix gap collapsing during PCDATA parsing for fragment mode.Arseny Kapoulkine
git-svn-id: https://pugixml.googlecode.com/svn/trunk@985 99668b35-9821-0410-8761-19e4c4f06640