Readd check for memory buffers without a read callback to avoid
XML_MAX_LOOKUP_LIMIT errors if users provide a custom input buffer.
Regressed with commit 834b8123 and later changes.
xmlParserInputBufferCreateMem must make a copy of the buffer.
This fixes a regression from 2.11 which could cause reads from freed
memory depending on the use case.
Undeprecate xmlParserInputBufferCreateStatic which can avoid copying
the whole buffer.
We now follow a laissez-faire approach when handling malloc failures and
removed many checks whether the parser was stopped by such an error.
This means the parser input must not be truncated to avoid out-of-bounds
reads.
Short-lived regression.
Fix many places where malloc failures aren't reported.
Make xmlErrMemory public. This is useful for custom external entity
loaders.
Introduce new API function xmlSwitchEncodingName.
Change the way how we store whether the the parser is stopped. This used
to be signaled by setting ctxt->instate to XML_PARSER_EOF which was
misdesigned and error-prone. Set ctxt->disableSAX to 2 instead and
introduce a macro PARSER_STOPPED. Also stop to remove parser inputs in
xmlHaltParser. This allows to remove many checks of ctxt->instate.
Introduce xmlErrParser to handle errors if a parser context is
available.
Fix many places where malloc failures aren't reported.
Rework XPath object cache to store free objects in a linked list to
avoid allocating an additional array. Remove some unneeded object pools.
Don't ignore malloc failures in xmlRaiseError and xmlCopyError.
Don't print filename if context has no input.
Introduce xmlVRaiseError taking a va_list.
Fix many places where malloc failures aren't reported.
Make some API function return an error code. Changing the return type
from void to int is technically an ABI break but should be safe on most
platforms.
- xmlNodeSetContent
- xmlNodeSetContentLen
- xmlNodeAddContent
- xmlNodeAddContentLen
- xmlNodeSetBase
Introduce new API functions that return a separate error code if a
memory allocation fails.
- xmlNodeGetAttrValue
- xmlNodeGetBaseSafe
- xmlGetNsListSafe
Introduce private functions xmlTreeEnsureXMLDecl and xmlSplitQName4.
Don't report low-level errors to the global error handler.
Fix tree
Introduce xmlGetNsListSafe
Fix tree
Introduce a new API function xmlAddIDSafe that returns a separate error
code if a memory allocation fails.
Store a pointer to the ID struct in xmlAttr so attributes can be
freed without allocating memory. It's impossible to report malloc
failures in deallocation code.
Fix places where malloc failures aren't reported.
Introduce a new API function xmlPatternCompileSafe that returns a
separate error code if a memory allocation fails.
Fix places where malloc failures aren't reported.
Introduce new API function xmlAddEntity that returns separate error
codes.
Don't invoke global error handler for low-level errors which should be
handled by higher layers.
Invalid redelcaration warnings will be fixed later.
Introduce new API functions that return a separate error code if a
memory allocation fails.
- xmlOpenCharEncodingHandler
- xmlLookupCharEncodingHandler
Fix a few places where malloc failures weren't reported.
Fix many places where malloc failures weren't reported, for example
after calling xmlStrdup.
Introduce new public API functions that return a separate error code if
a memory allocation fails:
- xmlParseURISafe
- xmlBuildURISafe
- xmlBuildRelativeURISafe
Update the fuzzer to check whether malloc failures are reported.
This regressed in commit e0dd330b.
Also fixes a long-standing issue where namespaces from default
attributes weren't added if they match an existing namespace.
Fixes#643.
Setting these deprecated globals hasn't had an effect for a long time.
Make them constants. This reduces the size of per-thread storage from
~700 to ~250 bytes.
The global struct is quite large (~700 bytes on 64-bit systems which
will be allocated for each thread whether it uses libxml2 or not) and
already close to the total size limit on some platforms.
Disable compiler TLS by default.
Set the dictionary for newDoc in xmlParseBalancedChunkMemoryRecover.
This is a long-standing bug which was masked by
- xmlParseBalancedChunkMemoryRecover changing the document of the root
node. This is a really bad idea, resulting in a mismatch between
ctxt->myDoc and ctxt->node->doc.
- SAX2.c preferring ctxt->node->doc over ctxt->myDoc until commit
a31e1b06.
Fixes#641.
The long-term plan is to remove the built-in HTTP client. There are
still a few downstream projects that use libxml2's client for other
purposes. These users should get deprecation warnings now.
