1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
|
New in 1.1.9:
* SAV reader: Improved support for Asian code pages #263
* SAV reader: Improved support for Very Long String records #287
* SAS reader: Improved support for RLE decompression #245 #253
* SAS reader: Support 16MiB page sizes #286
* SAS catalog reader: Fix bugs reading big-endian files #293
* SAS catalog reader: Allow formats with no labels #290
* SAS catalog reader: Check for long names in 64-bit files #291
* Improved compatibility with -Wstrict-prototypes #295
* Replace sprintf with snprintf #292
New in 1.1.8:
* XPT reader/writer: Improved support for format strings #257 #258
* DTA writer: Fix off-by-one error in v,o indexing for string refs #270
* SAV/DTA writers: Improved checking of non-ASCII characters #256
* SAS7BDAT reader: Fix use-after-free error #273
* Build: Link to libm on GNU systems #255
* SAS commands: Support more syntax
* SPSS commands: Make file names optional
New in 1.1.7:
* SAV writer: Permit missing ranges for integer variables #251
* SAV writer: Fix format-truncation warning #247
* DTA reader/writer: Improved support for string refs on older platforms
* DTA reader: Fix incorrect ordering in StrL comparison functions #248
* SAS7BDAT writer: Improved support for tagged missing values on big-endian platforms
* SAS readers: Allow header sizes up to 16MiB #249
* Visual Studio: Fix paths to make them consistent between x86 and x64 archs #242
New in 1.1.6:
* Migrate to GitHub Actions
* Regenerate parsers with Ragel 7 and update build script
* SAS7BDAT reader: Improved large file support on Windows #226
* SAV reader: Skip null bytes in UTF-8 data https://github.com/tidyverse/haven/issues/560
* SAV reader: Fix hang (oss-fuzz/23485)
* DTA reader: Disallow str0 type
* DTA reader: Fix encoding error when garbage values are present beyond the end of a string
* Command file readers: Fix integer overflow (oss-fuzz/15778)
* `extract_metadata`: Implement duration support #223 (thanks to @basgys)
* Support for SAS files created with SAS Visual Forecaster #232
* Report format widths for date/time SAS formats #233
* Document the meaning of a -1 return value from `readstat_get_row_count` #234
* Fix SAS file creation / modification times on Windows #238 #240
New in 1.1.5:
* Support for building with MSVC #214 (thanks to @zebrys and @jonathon-love)
* CLI tools: Support non-ASCII file paths on Windows #200 #216 (thanks to @zebrys)
* DTA reader: Ignore bad timestamps
* DTA writer: Fix memory leak
* DTA writer: Improved support for empty value labels #219
* POR reader: Improved support for date/time formats #160
* SAS7BDAT reader: Added support for reading the dataset label #180 #213 (thanks to @reikoch)
* SAS7BDAT reader: Improved detection of compressed files
* SAS7BDAT reader: Improved bounds checking OSS-Fuzz/28312
* SAS7BDAT reader: Support for more character encodings
* SAV reader: Tolerate illegal lowercase variable names #217
* SAV reader: Better support for non-UTF-8 variable names
* SAV reader: Fix format widths for very long strings https://github.com/Roche/pyreadstat/issues/77
* SAV reader: Fix undefined behavior with negative row counts OSS-Fuzz/23423
New in 1.1.4:
* SAS7BDAT reader: Add support for binary-compressed files #21
* XPT v8 writer: Improve compatibility with SAS #207 (thanks to @reikoch)
* XPT reader: Fix reading of long variable names #208 (thanks to @reikoch)
* SAS readers: Support for more character encodings
* SAV reader: Clients sometimes received truncated UTF-8 strings
* SPSS writers: Improve compatibility with PSPP with DATETIME fields #211
* All formats: Improved support for setting / getting the `display_width` #210
New in 1.1.3:
* Fix warnings when compiling with GCC 10 #202
* SAS RLE compressor: Fixes for large files #201
* SAV reader: Improved support for UTF-8 column names #206
* SAV reader: Return a better error message if the magic number doesn't match
* SAV reader: Support for dash-separated timestamps
New in 1.1.2:
* DTA reader: support for Spanish-locale timestamps
* SAS reader: support for "any" encoding tidyverse/haven#482
* CLI tool: Allow uppercase filename extensions
* Improved support for reading SPSS and SAS command files
* Improved support for reading POR files with format widths >100
* Improved support for reading SAV files containing space-padded timestamps #197
* Improved support for writing SAV files with a large number of variables #199
* Improved support for reading SAS7BDAT files created by Stat/Transfer #198
* Fix several integer overflows and undefined values #192 #193 #194 #195 #196
New in 1.1.1:
* Support row limits in the plain-text parsers
* SAV reader: Allow spaces in timestamp strings
* README: Fix Windows / pacman instructions #189
* Fix </sortlist> errors opening files in Stata 15 (tidyverse/haven#461)
New in 1.1:
* New function: readstat_set_row_offset (#185). Thanks to @mikmart
* Fix segfault when localtime fails on Windows
* Fix implicit float conversion warning (oss-fuzz/16372)
* New error code: READSTAT_ERROR_BAD_TIMESTAMP_VALUE
* Renamed error code: READSTAT_ERROR_BAD_TIMESTAMP => READSTAT_ERROR_BAD_TIMESTAMP_STRING
New in 1.0.2:
* Compilation: Fix -Wstringop-truncation warnings on GCC 8.2 and later (#151)
* SPSS command parser: Fix signed integer overflow (oss-fuzz/15049)
* POR parser: Use doubles internally to prevent integer overflows with very large exponents (#182)
New in 1.0.1:
* SAV writer: Validate variable names
* Fix a buffer overflow reading SPSS commands (oss-fuzz/15050)
* New error code READSTAT_ERROR_NAME_IS_ZERO_LENGTH when a blank variable name is provided
* New fuzzing dictionary files in fuzz/dict for parsing plain-text file formats
* Move corpus files from corpus to fuzz/corpus
|
