|
29beef65
|
2024-01-02T21:50:38
|
|
parser: Pop inputs if parsing DTD failed
This should provide some statistics in ctxt->sizeentcopy even in the
error or recovery case.
|
|
|
02a2038d
|
2024-01-10T14:17:49
|
|
parser: Handle NOCDATA properly when expanding entities
Short-lived regression from e1153832.
|
|
|
fd801845
|
2024-01-07T15:19:58
|
|
fuzz: Cap URL size
Cap URL size to avoid quadratic behavior when generating error messages.
|
|
|
83c1ae13
|
2024-01-07T15:40:23
|
|
fuzz: Add missing include
Fix build failure.
|
|
|
e1153832
|
2024-01-07T01:29:37
|
|
parser: Fix quadratic behavior when copying entities
Process the first and last text node with the SAX handler to make the
text merging optimization kick in.
Fixes #657.
|
|
|
d2b55a7a
|
2024-01-05T20:31:10
|
|
writer: Implement xmlTextWriterClose
This function can be used to make sure that closing the output stream
succeeded.
Fixes #513.
|
|
|
02cc5c36
|
2024-01-05T04:17:14
|
|
parser: Add XML_PARSE_NO_XXE parser option
|
|
|
12f0bb94
|
2024-01-05T01:14:28
|
|
parser: Synchronize more options
|
|
|
3efbe916
|
2024-01-05T00:11:29
|
|
parser: Mark 'token' member as unused in xmlParserCtxt
|
|
|
b82fd81d
|
2024-01-04T23:25:06
|
|
parser: Rework xmlCtxtParseDocument
Make xmlCtxtParseDocument take a parser input which can be popped after
parsing.
|
|
|
f237e5b9
|
2024-01-05T15:40:23
|
|
parser: Avoid duplicate namespace errors
Don't report an extra attribute uniqueness error if a namespace is
undeclared. This matches old behavior.
|
|
|
c2b3294f
|
2024-01-04T21:20:51
|
|
fuzz: Abort on invalid UTF-8
The parser should never generate invalid UTF-8 these days even in
recovery mode.
|
|
df098e3b
|
2023-12-22T12:02:08
|
|
Set LIBXML2_FOUND if it has been properly configured
|
|
|
d7d300ba
|
2024-01-04T17:50:11
|
|
parser: Remove remnants of runtime debugging feature
Apparently, this feature was remove long ago.
Fixes #651.
|
|
|
8c5848bd
|
2024-01-04T17:14:31
|
|
parser: Make xmlParseContent more useful
This is an internal function which isn't really usable without some
hacks. See WebKit/Chromium trying to recreate the effects of
xmlDetectSAX2 manually, for example.
Make xmlParseContent perform late initialization and check whether the
content was fully parsed.
Also rename xmlDetectSAX2 and document why it's needed.
|
|
|
65c65b65
|
2024-01-04T13:59:23
|
|
tests: Move away from global error handlers
|
|
|
07c05546
|
2024-01-04T02:48:02
|
|
error: Make xmlFormatError public
This is a useful function to get a verbose error report.
Allows to remove duplicated code from runtest.c. Also reactivate check
for schema parser failures.
|
|
|
d0eb5a7e
|
2024-01-03T18:12:29
|
|
parser: Remove xmlErrEncodingInt
Convert the last user to xmlFatalErr.
|
|
|
f30b9b23
|
2024-01-03T18:11:44
|
|
fuzz: Add assertion in xmlCopyCharMultibyte
This is an internal function that should never receive out-of-range
codepoints.
|
|
|
a7356dfe
|
2024-01-03T18:02:46
|
|
parser: Clear invalid entity content
This was removed in earlier commits, but we really want to make sure
that entity content is syntactically valid.
|
|
|
30d83977
|
2024-01-04T15:18:14
|
|
fuzz: Disable catalogs
The catalogs API doesn't report OOM errors. It's basically impossible
to use it safely in its current form.
|
|
|
ca5965d5
|
2024-01-02T21:49:43
|
|
save: Report more malloc failures
|
|
|
2c9cd0b6
|
2024-01-02T18:51:24
|
|
fuzz: Abort on internal errors
|
|
|
661ef936
|
2024-01-02T18:50:59
|
|
valid: Fix some error codes
|
|
|
0821efc8
|
2024-01-02T18:33:57
|
|
encoding: Check whether encoding handlers support input/output
The "HTML" encoding handler doesn't support input which could lead to a
wrong error report.
|
|
|
85f99023
|
2024-01-02T17:52:43
|
|
parser: Fix buffer size checks
Don't test size of remaining data. This causes false positives with
memory buffers.
Also impose XML_MAX_HUGE_LENGTH limit when parsing with XML_PARSE_HUGE.
|
|
|
e8fb3d63
|
2024-01-02T17:45:54
|
|
parser: Convert some "internal errors" to meaningful codes
|
|
|
9912c369
|
2024-01-02T17:23:59
|
|
SAX2: Enforce size limit in xmlSAX2Text with XML_PARSE_HUGE
|
|
|
5cb4b05c
|
2024-01-02T17:16:22
|
|
parser: Lower maximum entity nesting depth
Limit entity nesting depth to 20 or 40 with XML_PARSE_HUGE.
Change error code to XML_ERR_RESOURCE_LIMIT.
|
|
|
a2cc7f5f
|
2024-01-02T17:02:21
|
|
parser: Set depth limit to 2048 with XML_PARSE_HUGE
Deeply nested documents can cause performance problems, so the nesting
depth should always be limited to a reasonable value.
Also remove the global xmlParserMaxDepth setting which isn't thread-safe
and seems unused.
|
|
|
875bb084
|
2023-09-07T03:25:45
|
|
parser: Implement xmlCtxtSetOptions
Surprisingly, some options can only be enabled with xmlCtxtUseOptions
and it's impossible to unset them. Add a new API function
xmlCtxtSetOptions which sets or clears all options.
Finally document all parser options.
Make sure to synchronize option bits and struct members.
|
|
|
33ec407a
|
2023-09-07T03:33:09
|
|
parser: Always prefer option members over bitmask
If an option has an extra member in xmlParserCtxt, it takes precedence
over the value from the options bitmask. Fix a few places where this was
ignored.
|
|
|
22fd571f
|
2023-09-06T22:15:20
|
|
parser: Don't modify SAX2 handler if XML_PARSE_SAX1 is set
It's a bad idea to modify members of the SAX handler struct for option
state management. Ideally, ctxt->options should be the preferred source
of truth.
|
|
|
37c6618b
|
2023-12-30T02:50:34
|
|
parser: Rework parsing of attribute and entity values
Don't use a separate function to handle "complex" attributes. Validate
UTF-8 byte sequences without decoding. This should improve performance
considerably when parsing multi-byte UTF-8 sequences.
Use a string buffer to avoid unnecessary allocations and copying when
expanding entities.
Normalize attribute values in a single pass while expanding entities.
Be more lenient in recovery mode.
If no entity substitution was requested, validate entities without
expanding. Fixes #596.
Also fixes #655.
|
|
|
4dcc2d74
|
2024-01-02T14:04:44
|
|
save: Output U+FFFD replacement characters
This degrades more gracefully and helps to diagnose errors.
We stop raising errors for now, since there's no way to report malloc
failures during error handling yet.
|
|
|
2b79f106
|
2023-12-29T21:07:04
|
|
parser: Simplify entity size accounting
|
|
|
08d9b258
|
2023-12-29T15:20:56
|
|
parser: Support namespace scope in NsData struct
The previous approach of recreating the NsData struct was flawed.
|
|
|
5de48d12
|
2023-12-29T14:41:40
|
|
parser: Simplify error handling when parsing entities
|
|
|
f0dc52d0
|
2023-12-29T06:00:20
|
|
parser: Move cleanup of element stacks to xmlParseContent
|
|
|
a1ed589b
|
2023-12-29T23:12:06
|
|
parser: Avoid unwanted expansion of parameter entities
Remove PE handling from xmlSkipBlankChars and add a separate version
that handles PEs. Only call xmlSkipBlankCharsPE when parsing DTD
constructs. This should make sure that PEs don't get expanded
accidentally, for example in text declarations.
|
|
|
16b0dbc1
|
2023-12-29T18:47:30
|
|
parser: Fix XML_ERR_UNSUPPORTED_ENCODING errors
Commit 45157261 added the check in the wrong place.
Also allow unsupported encoding in xmlNewInputInternal.
Fixes #654.
|
|
|
e45a4d71
|
2023-12-29T00:00:21
|
|
io: Always forward IO errors to global handler
The HTTP module raises errors without context. This won't be fixed,
so send them to the global error handler.
|
|
|
a73483ed
|
2023-12-29T00:22:02
|
|
parser: Remove extraneous error message
This is not an "internal error" but some other error reported elsewhere.
|
|
|
7e0bbbc1
|
2023-12-27T18:33:30
|
|
parser: New input API
Provide a new set of functions to create xmlParserInputs. These can be
used for the document entity or from external entity loaders.
- Don't require xmlParserInputBuffer.
- All functions take a base URI.
- All functions take an encoding as string.
- xmlNewInputURL also takes a public ID.
- xmlNewInputMemory takes a size_t.
- Optimization hints for memory buffers.
Improve documentation.
Only call xmlInitParser before allocating a new parser context.
Call xmlCtxtUseOptions as early as possible.
|
|
|
45157261
|
2023-12-27T21:30:13
|
|
parser: Downgrade XML_ERR_UNSUPPORTED_ENCODING to warning
If the actual encoding is UTF-8 or ASCII, we don't want to fail.
|
|
|
24b7144f
|
2023-12-27T15:50:58
|
|
parser: More refactoring of entity parsing
Remove xmlCreateEntityParserCtxtInternal.
Rework xmlNewEntityInputStream.
|
|
|
d3ceea0b
|
2023-12-27T15:18:09
|
|
parser: Fix encoding handling in xmlParserInputBufferCreateIO
Don't pass encoding to xmlParserInputBufferCreateIO but use
xmlSwitchEncoding to make sure that the encoding sticks.
|
|
|
d025cfbb
|
2023-12-27T03:53:24
|
|
parser: Always copy content from entity to target.
Make sure that references from IDs are updated.
Note that if there are IDs with the same value in a document, the last
one will now be returned. IDs should be unique, but maybe this should be
addressed.
|
|
|
6337ff79
|
2023-12-27T03:29:13
|
|
parser: Simplify control flow in xmlParseReference
|
|
|
579186f2
|
2023-12-27T03:03:26
|
|
parser: Remove xmlSetEntityReferenceFunc feature
This has been deprecated for a long time.
|
|
|
b848338c
|
2023-12-27T01:46:40
|
|
parser: More refactoring of entity loading
This sets input->entity also for general entities.
|
|
|
4ecc85d2
|
2023-12-27T00:44:16
|
|
parser: Push general entity input streams on the stack
This allows the error handler to give more context.
|
|
|
a5dcf0f4
|
2023-12-26T03:27:23
|
|
parser: Mark more parser context members as unused
|
|
|
6a9a88a1
|
2023-12-26T03:13:05
|
|
parser: Move progressive flag into input struct
|
|
|
4f14fe9c
|
2023-12-26T02:44:38
|
|
parser: Remove remaining ctxt->instate checks
Now ctxt->instate is only used for push parser states.
|
|
|
d944a415
|
2023-12-26T02:10:35
|
|
parser: Fix in-parameter-entity and in-external-dtd checks
Use in ctxt->input->entity instead of ctxt->inputNr to determine whether
we are inside a parameter entity.
Stop using ctxt->external to check whether we're in an external DTD.
This is signaled by ctxt->inSubset == 2.
|
|
|
477a7ed8
|
2023-12-28T19:06:32
|
|
html: Abort earlier on fatal errors
|
|
|
5f319304
|
2023-12-28T19:05:51
|
|
SAX2: Fix error code
Today I learned that the TSCII character encoding [1] can blow up the
size of text 12 times when converted to UTF-8:
$ printf '\x82' |iconv -f TSCII -t UTF-8 |hexdump -C
00000000 e0 ae b8 e0 af 8d e0 ae b0 e0 af 80
0000000c
[1] https://en.wikipedia.org/wiki/Tamil_Script_Code_for_Information_Interchange
|
|
|
ab631971
|
2023-12-28T17:07:03
|
|
uri: Keep fragment intact when resolving filesystem paths
|
|
|
b8313b58
|
2023-12-26T21:59:08
|
|
xpath: Rewrite substring-before and substring-after
Don't use buffers. Check malloc failures.
|
|
|
3874e5d0
|
2023-12-26T01:42:23
|
|
tests: Remove unneeded error formatting code
|
|
|
2a2fbe1e
|
2023-12-28T16:42:03
|
|
xinclude: Only set xml:base if necessary
|
|
|
8a685a3d
|
2023-12-26T00:42:22
|
|
xinclude: Allow empty nodesets
There's no reason to treat an empty nodeset as error.
|
|
|
f3fa34dc
|
2023-12-26T22:37:26
|
|
parser: Fix general entity parsing
Clear namespace database.
Ignore non-fatal errors.
|
|
|
ecfbcc8a
|
2023-12-25T04:33:00
|
|
parser: Rework general entity parsing
Don't create a new parser context but reuse the existing one.
This exposes bug #601 in a more obvious way.
|
|
|
c2ef78f7
|
2023-12-24T23:56:57
|
|
io: Fix close error handling
There's no way to report error codes from closing an output buffer yet.
|
|
|
6d27c549
|
2023-12-24T17:59:02
|
|
io: Fix read/write error handling
Handle short reads/writes from fd. Fix stdio error handling.
|
|
|
0bef93bf
|
2023-12-23T04:03:41
|
|
io: More refactoring and unescaping fixes
Merge Windows wrappers into relevant functions.
Remove more unnecessary unescaping.
Merge *OpenW into *Open functions.
Use unbuffered IO for output.
|
|
|
331dcd62
|
2023-12-23T01:40:54
|
|
error: Reenable full error reports to default handler
This should make console output include some information about nodes
again. Note that this extra information must be disabled if a custom
generic error handler was set. Many downstream test suites rely on this
behavior.
|
|
|
c1bddd4c
|
2023-12-23T01:09:17
|
|
parser: Mark 'length' member of xmlParserInput as unused
|
|
|
955c177f
|
2023-12-23T00:58:36
|
|
parser: Stop using 'directory' struct member
This was only used as a pointless fallback for URI resolution.
|
|
|
60841beb
|
2023-12-25T18:31:22
|
|
parser: Make XML_IO_NETWORK_ATTEMPT behave as before
Always reported to generic error, not to parser context for backward
compatibility. Several downstream test suites rely on this behavior.
|
|
|
a2693410
|
2023-12-23T00:35:30
|
|
io: Move some code from xmlIO.c to parserInternals.c
Move everything related to parser contexts to parserInternals.c.
|
|
|
8ab1b122
|
2023-12-23T00:00:15
|
|
Fix filename and URI handling
Many strings are passed to the library that could be either URIs or
filesystem paths. We now assume that strings are a URI if they contain
the substring "://". This means that they have a scheme and an
authority. Otherwise, URI resolution wouldn't make much sense.
Fix xmlBuildURI to work with filesystem paths. If the base URI doesn't
contain "://" it is treated as filename. The resolved URI is unescaped,
appended and the result is normalized. Rewrite xmlNormalizePath to
handle Windows quirks.
All special handling for Windows paths is removed in xmlCanonicPath.
If the path looks like an URI, only escape characters allowed in Legacy
Extended IRIs.
Make xmlPathToURI only call xmlCanonicPath. Theh additional round-trip
through URI parser and serializer seems useless.
Add a helper function xmlConvertUriToPath in xmlIO.c which checks for
file URIs and unescapes them.
Always process strings with xmlCanonicPath in xmlLoadExternalEntity.
This should be harmless now.
Should help with #334, #387, #611.
|
|
|
28913232
|
2023-12-22T23:58:43
|
|
uri: Clean up special parsing modes
Add function to handle unreserved check. Give flags meaningful names.
Add support to allow ucschars from Legacy Extended IRIs.
|
|
|
6e3a2ac6
|
2023-12-22T21:38:50
|
|
xinclude: Rework xml:base fixup
The xml:base fixup was broken in more complex cases.
Also avoid parsing and building the included URI multiple times.
|
|
|
35a4bc50
|
2023-12-22T15:14:19
|
|
xinclude: Report to xmlGenericError
|
|
|
e8de3401
|
2023-12-22T02:57:19
|
|
parser: Also set document properties when push parsing
Add new function xmlFinishDocument which invokes the endDocument SAX
handler and sets the document's properties.
|
|
|
c73de050
|
2023-12-23T04:50:47
|
|
include: Move non-generated parts from xmlversion.h.in
xmlexports.h originally only included symbol visibility macros but it's
a good place for other macros as well.
|
|
|
a18d9416
|
2023-12-21T18:39:44
|
|
Update NEWS
|
|
|
229e5ff7
|
2023-12-21T18:09:42
|
|
io: Remove support for HTTP POST
This feature is unlikely to be used these days.
|
|
|
9c2c87b5
|
2023-12-24T15:33:12
|
|
dict: Move local RNG state to global state
Don't use TLS variables directly.
|
|
|
2e9e758d
|
2023-12-24T14:27:46
|
|
dict: Get random seed from system PRNG
|
|
|
c49572e5
|
2023-12-23T15:03:22
|
|
malloc-fail: Fix erroneous report in xmlStringGetNodeList
The parser can produce invalid attribute content in recovery mode.
Unless this is fixed, xmlStringGetNodeList should ignore such errors
silently.
|
|
|
c8f1f4a2
|
2023-12-21T17:30:38
|
|
doc: Improve documentation of error handlers
|
|
|
882b3a80
|
2023-12-21T15:34:24
|
|
runtest: Fix return code in rngTest
|
|
|
f0df3e6d
|
2023-12-21T14:35:18
|
|
tests: Try to fix RelaxNG test cases
These were added recently in ea695ac0 and 8074b881 but were a total mess
of symbolic links and apparently mixed up files.
Symbolic links don't work on Windows.
Try to salvage one of the tests.
|
|
|
8cd56317
|
2023-12-21T02:32:01
|
|
html: Don't close fd in htmlCtxtReadFd
Long-standing bug. The XML fix from 2003 was never ported to the HTML
parser. htmlReadFd was fixed with fe6890e2.
|
|
|
0a658c0f
|
2023-12-20T23:53:19
|
|
io: Don't use "-" to read from stdin
To implement this feature on such a low level is a disaster waiting to
happen. Remove these checks from the IO code and move them to xmllint.
Note that the serialization API will still treat "-" as stdout.
|
|
|
c9a46a91
|
2023-12-20T20:11:09
|
|
io: Rework initialization
|
|
|
b75fc1ab
|
2023-12-20T20:01:19
|
|
io: Rearrange code
|
|
|
13043691
|
2023-12-20T00:33:34
|
|
parser: Rename xmlErrParser to xmlCtxtErr
|
|
|
8d0aaf4b
|
2023-12-19T20:47:36
|
|
parser: Remove xmlErrEncoding
Use xmlFatalErr or xmlCtxtErrIO.
|
|
|
9fbe46ba
|
2023-12-19T20:10:10
|
|
io: Consolidate error messages
|
|
|
23345a1c
|
2023-12-19T19:52:28
|
|
io: Report IO errors through xmlCtxtErrIO
This is also a new public API function to be used in external entity
loaders.
|
|
|
e62b0dbd
|
2023-12-19T19:47:07
|
|
xzlib: Fix harmless unsigned integer overflow
|
|
|
1ef35663
|
2023-12-19T19:36:35
|
|
io: Always use unbuffered input
Before, we often used unbuffered input via the lzma or gzip handlers,
more or less inadvertently.
Change the default file handlers from buffered (stdc FILE) to unbuffered
(POSIX fds).
|
|
|
7e14c05d
|
2023-12-19T17:05:08
|
|
io: Fix detection of compressed streams
Make sure that we don't try to open uncompressed streams with a
compression handler in copying mode.
|
|
|
7e511f35
|
2023-12-19T15:41:37
|
|
io: Pass error codes from xmlFileOpenReal to xmlNewInputFromFile
This allows to report the reason why opening a file failed to the parser
context and improve error messages. Now we can also remove the stat call
before opening a file.
|
|
|
b2dbcc43
|
2023-12-19T13:33:59
|
|
io: Rework default callbacks
Register a dummy callback struct for default callbacks. Handle them in a
separate function which will later allow to return meaningful error
codes.
|