In fpm-log, possible buffer overflow. Check for length is done at
the beginning of the loop, so is not done when overflow occurs
on the last loop (len = 1024 or 1025). (ack from fat).
This issue where found from by static code analysis tool and, so,
I can't provide any reproducer.
* 'PHP-5.4' of git.php.net:php-src:
BFN #61272
fix bug #61272 ob_start callback get passed empty string
Fixed bug #63590 Fileinfo delivers ifferent results in TS and NTS under Windows
This test isn't supposed to pass on windows
Fixed Bug #63614 (Fatal error on Reflection)
Possible NULL dereference when trying to delete the single item
of a list (ack from fat).
This issues where found from by static code analysis tool and,
so, I can't provide any reproducer.
* 'PHP-5.3' of git.php.net:php-src:
Fixed bug #63590 Fileinfo delivers ifferent results in TS and NTS under Windows
This test isn't supposed to pass on windows
Fixed Bug #63614 (Fatal error on Reflection)
* PHP-5.3:
Fixed bug #63590 Fileinfo delivers ifferent results in TS and NTS under Windows
Conflicts:
NEWS
ext/fileinfo/libmagic.patch
ext/fileinfo/tests/finfo_open_error-win32.phpt
Following comment from Yoram "The patch looks fine, except of testing
for true value of utf16 in each iteration."
Also fix the length computation during check phase.
Json use an utf8 parser from a third party library, switch to
our implementation of php_next_utf8_char.
This also helps on solving #63520. All the unit tests succeed.
Our implementation also seems a little faster.
json.dsp need to be regenerated.
In pdo_sqlite, getColumnMeta should return the table name.
With bundled SQLite library, -DSQLITE_ENABLE_COLUMN_METADATA=1 (other_flags)
enable the feature in SQLite (function sqlite3_column_table_name).
With system SQLite library, this fix detects if sqlite3_column_table_name is
available and add SQLITE_ENABLE_COLUMN_METADATA macro to get the same behavior.
Test already exists: ext/pdo_sqlite/tests/bug_42589.phpt