drobilla.net - RDF

Sratom 0.6.22

2026-02-10T18:45:00-05:00

Sratom 0.6.22 has been released. Sratom is a small library for serializing LV2 atoms. Sratom reads/writes atoms from/to RDF, allowing them to be converted between binary and text or stored in a model.

Changes:

Add clang nullability annotations
Address new warnings in clang and clang-tidy 21
Fix documentation build without sphinx_lv2_theme
Gracefully handle reading vectors with missing childType properties
Gracefully handle writing vectors with zero childSize properties
Improve error handling

Sord 0.16.22

2026-02-10T18:43:00-05:00

Sord 0.16.22 has been released. Sord is a lightweight C library for storing RDF statements in memory.

Changes:

Add clang nullability annotations
Address new warnings in clang and clang-tidy 21
Make more API functions tolerate NULL

Serd 0.32.8

2026-02-10T18:02:00-05:00

Serd 0.32.8 has been released. Serd is a lightweight C library for working with RDF data.

Changes:

Avoid use of snprintf() when writing
Fix unbounded reader stack growth when reading NQuads
Improve reader error handling
Relax nullability of SerdNode buffer field
Release reader stack memory on read error
Strengthen handling of corrupt UTF-8 input
Support writing all escapes in Turtle and TriG prefixed names

Sratom 0.6.20

2025-11-12T19:26:00-05:00

Sratom 0.6.20 has been released. Sratom is a small library for serializing LV2 atoms. Sratom reads/writes atoms from/to RDF, allowing them to be converted between binary and text or stored in a model.

Changes:

Avoid over-use of yielding meson options
Avoid use of scanf when reading MIDI events
Fix lint checks
Fix potential memory error when writing ambiguous relative paths
Improve code quality

Sord 0.16.20

2025-11-12T19:12:00-05:00

Sord 0.16.20 has been released. Sord is a lightweight C library for storing RDF statements in memory.

Changes:

Add header warnings test
Add missing const qualifiers to API
Avoid over-use of yielding meson options
Update man pages

Serd 0.32.6

2025-11-12T18:54:00-05:00

Serd 0.32.6 has been released. Serd is a lightweight C library for working with RDF data.

Changes:

Avoid over-use of yielding meson options
Drop graphs when writing Turtle output
Eliminate recursion in the writer
Fix handling of bad predicates in anonymous blank nodes
Fix handling of some invalid EOF cases in lax mode
Fix indentation of named objects after anonymous objects
Fix indentation when ending anonymous nodes with many objects
Fix invalid characters in error messages
Fix reading numbers with no space before the final dot
Fix reading prefix names that start with "true." or "false."
Refuse to write incoherent statements
Remove project and version number from man page OS field
Write a blank line between statements and Turtle/TriG directives

Lilv “Object Index” Option Revised

2025-11-12T13:58:00-05:00

My previous post, “Improving LV2 Discovery Performance”, described adding an option to lilv that controls whether an additional index is maintained for loaded LV2 data. While testing those changes for release, I discovered a few things (like lv2lint) that break without the index, which convinced me that disabling it by default wasn't the right decision.

So, I'm instead enabling it by default, which avoids this “break by default” situation. The downside is that hosts will need to add code in order to enjoy the performance improvements. To make this change as simple as possible, I also made lilv_world_set_option accept a null value to disable boolean options, so most hosts can simply add this line:

lilv_world_set_option(world, LILV_OPTION_OBJECT_INDEX, NULL);

This requires the new version, but the same change can be made without bumping the required version by avoiding the niceties:

LilvNode* false_node = lilv_new_bool(world, false);

lilv_world_set_option(
  world,
  "http://drobilla.net/ns/lilv#object-index",
  false_node);

lilv_node_free(false_node);

This will log a warning about the unsupported option on older versions, but otherwise have no effect.

Improving LV2 Discovery Performance

2025-09-29T10:25:00-04:00

LV2 plugins (and other data, like presets) are discovered by parsing Turtle files installed in various bundle directories. If the host only needs to know which things are installed, then only manifest.ttl files need to be loaded. That takes only a fraction of a second (60 ms on this machine, which has 897 plugins installed), but if the host needs more information (like the plugin's label or what ports it has), then the data files need to be loaded as well, which can take several seconds. The parser (from serd) is very fast, so the overwhelming majority of this time is spent inserting data into a model.

I was recently doing some work on lilv related to discovery, and got sidetracked into investigating how much of this overhead could be eliminated. Quite a bit, as it turns out, but before explaining how, I'll give a brief summary of the relevant fundamentals so people who don't spend their days in the triple mines can understand what I'm talking about.

LV2 data is written in the Turtle syntax, which can have a lot of structure, but is ultimately just a shorthand for data that is simply a “flat” series of triples. For example, the abbreviated Turtle

<http://example.org/amp>
    a lv2:Plugin ;
    doap:name "Amplifier" .

represents the two triples

<http://example.org/amp> rdf:type  lv2:Plugin .
<http://example.org/amp> doap:name "Amplifier" .

which describe a plugin with the name “Amplifier”.

Two implications of this format are important to understand here:

All data is a set of triples which can be stored in some lexicographical order to be quickly searchable (for example, a set ordered subject-first can be used to quickly find triples that start with a given subject).
All data can be streamed via a simple “flat” interface (for example a function with three parameters), and is trivial to inspect and/or filter on the fly (much like line-based text in POSIX pipelines).

The data may not be stored as a “literal” ordered set of triples (they're actually quads in memory for one thing), but this simplified way of thinking about it is good enough for this explanation.

Which order is best depends on the query. For example (borrowing a bit of syntax from SPARQL where ?name represents a query variable), if the host asks something like “what's the name of this plugin” or

<http://example.org/amp> doap:name ?name .

then it needs to find triples that start with the given subject and predicate (<http://example.org/amp> and doap:name) to see what object (?name) they have. So, a subject-predicate-object or “SPO” index will work well. If, however, the host asks something like “which things are plugins” or

?plugin rdf:type lv2:Plugin .

then an SPO index doesn't help because there's no known subject to search for. Most queries are like this, with either a subject or object wildcard (querying relatively “fixed” predicates is rare), so lilv has always had both an SPO and OPS index to support them.

This is convenient, but twice the indexing means roughly twice the overhead. The SPO index is the natural order used in the syntax, and supports most code (which mainly looks up properties of known things), but the OPS index isn't used so much. Can it be removed entirely to speed things up? It is used for many important things, but conveniently, there's only a few fixed cases of those (like finding plugins as above). So, we can take advantage of the streaming nature of the data to record this information while it's being read, instead of inserting it into an index only to query it out later.

To do this in the implementation, I introduced the concept of a “skimmer”. A skimmer inserts triples into a model as usual, but first “skims” them and records items of interest for later use. For example, a skimmer checks whether a triple has rdf:type lv2:Plugin, and if so, records the subject in a result set that stores only plugin URIs. Some cases are a bit trickier, and actually pulling this off cleanly took quite an overhaul, but details aside, it turns out that this approach can be used to eliminate the OPS index entirely without losing any functionality.

How much of an improvement does this make? On this machine, using lv2ls -n as a crude benchmark, the time goes from 3.11 to 1.72 seconds, or about a 45% improvement. The memory consumption goes down a bit as well, from (even more crudely) a max-RSS of about 138 to 112 MiB, or about a 19% improvement (everything here is an average of three runs). The improvement will vary dramatically based on what's installed, for example in some earlier tests with a different configuration I saw drops from around 7 to 4 seconds, but in general, not bad!

Are there any downsides to this? Yes, primarily two:

As with any significant change, there is the possibility of regression. Discovery is a little more restricted than before, so although I've done my best to test things, there is a possibility of things not being discovered anymore. Usually the fix for this is adding information to the manifest that should have been there anyway.
Though the implementation itself doesn't make queries that require the OPS index, hosts themselves can since lilv provides generic query functions. To support this, there's a new boolean option, LILV_OPTION_OBJECT_INDEX, which can be used to enable or disable the OPS index. Since most hosts don't need it, I decided to disable this by default. That will be a regression for hosts that do, however, who need to explicitly opt in to the old behaviour by enabling this option. A warning is printed if a query will be slow because of the missing index, so it should at least be obvious if this happens.

These changes are now in the main branch in git, which will be released soon, probably as version 0.26.0.

Sratom 0.6.18

2025-01-19T13:49:00-05:00

Sratom 0.6.18 has been released. Sratom is a small library for serializing LV2 atoms. Sratom reads/writes atoms from/to RDF, allowing them to be converted between binary and text or stored in a model.

Changes:

Add lint option with project metadata and code formatting tests
Avoid snprintf when writing MIDI events
Enable clang nullability checks
Fix library current_version on MacOS

Sord 0.16.18

2025-01-19T13:46:00-05:00

Sord 0.16.18 has been released. Sord is a lightweight C library for storing RDF statements in memory.

Changes:

Add option to install tool man pages
Enable clang nullability checks
Fix build as C with MSVC
Fix library current_version on MacOS
Replace more platform-specific code with use of zix

Serd 0.32.4

2025-01-19T00:17:00-05:00

Serd 0.32.4 has been released. Serd is a lightweight C library for working with RDF data.

Changes:

Clean up enum declarations
Fix library current_version on MacOS
Fix overly permissive parsing of syntax names on the command line
Fix parsing NQuads lines with no space before the final dot
Fix reading chunks from files without trailing newlines
Fix rewriting special literals when datatype URIs are prefixed names
Gracefully handle errors while writing the end of anonymous nodes
Improve test suite coverage
Simplify some character classification and comparison code
Support reading lone lists in lax mode
Treat out of range unicode characters as errors
Write blank lines between graphs and statements in TriG

Serd 0.32.2

2024-01-23T19:32:00-05:00

Serd 0.32.2 has been released. Serd is a lightweight C library for working with RDF data.

Changes:

Enable clang nullability checks
Fix writing empty list objects within blank nodes

Sratom 0.6.16

2023-10-22T20:06:00-04:00

Sratom 0.6.16 has been released. Sratom is a small library for serializing LV2 atoms. Sratom reads/writes atoms from/to RDF, allowing them to be converted between binary and text or stored in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Clean up code
Constrain relative URI references to the base URI
Fix dependencies in pkg-config file
Override pkg-config dependency within meson
Remove junk files from documentation install
Replace duplicated dox_to_sphinx script with sphinxygen dependency

Sord 0.16.16

2023-10-22T19:59:00-04:00

Sord 0.16.16 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Allow SORD_API to be defined by the user
Fix command line help interface of sord_validate
Fix dependencies in pkg-config file
Override pkg-config dependency within meson
Port sord_validate to pcre2
Switch to external zix dependency

Serd 0.32.0

2023-10-22T19:36:00-04:00

Serd 0.32.0 has been released. Serd is a lightweight C library for working with RDF data.

Changes:

Add Windows path separator support to serd_node_new_file_uri()
Add long "help" and "version" options to serdi
Add options to disable html or singlehtml documentation
Add serd_reader_skip_until_byte() to public API
Allow SERD_API to be defined by the user
Avoid creating test files in the current directory
Avoid using ASCII grave as a quote
Check for POSIX features with the build system
Clean up and improve test suite
Clean up code
Fix crash when trying to read chunks without starting
Fix hang when skipping an error at EOF when lax parsing
Fix incorrect parsing of strange quote escape patterns
Fix possible hang when writing nested Turtle lists
Fix potential memory leaks when a write is aborted
Fix relative URI creation
Gracefully handle bad characters in Turtle blank node syntax
Gracefully handle bad characters in Turtle datatype syntax
Improve TriG pretty-printing and remove trailing newlines
Improve pretty-printing of lists and inline subjects
Improve serdi man page
Improve writer error handling
Make URI writing stricter by default
Make serd_reader_read_chunk() work with NQuads
Override pkg-config dependency within meson
Remove junk files from documentation install
Remove support for writing Turtle named inline nodes extension
Replace duplicated dox_to_sphinx script with sphinxygen dependency
Test header for warnings more strictly
Update standard test suites

Sratom 0.6.14

2022-09-09T17:29:00-04:00

Sratom 0.6.14 has been released. Sratom is a small library for serializing LV2 atoms. Sratom reads/writes atoms from/to RDF, allowing them to be converted between binary and text or stored in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Add project metadata
Adopt REUSE machine-readable licensing standard

Sord 0.16.14

2022-09-09T03:07:00-04:00

Sord 0.16.14 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Adopt REUSE machine-readable licensing standard
Allow programs to be used from subproject
Fix accidentally exposed internal zix symbols

Serd 0.30.16

2022-09-09T03:06:00-04:00

Serd 0.30.16 has been released. Serd is a lightweight C library for working with RDF data.

Changes:

Add SERD_STATIC to pkg-config Cflags for static-only builds
Adopt REUSE machine-readable licensing standard
Allow programs to be used from subproject
Fix spelling mistake in serdi man page

Sratom 0.6.12

2022-07-19T01:13:00-04:00

Sratom 0.6.12 has been released. Sratom is a small library for serializing LV2 atoms. Sratom reads/writes atoms from/to RDF, allowing them to be converted between binary and text or stored in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Switch to meson build system

Sord 0.16.12

2022-07-17T21:22:00-04:00

Sord 0.16.12 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Fix various warnings
Switch to meson build system

Serd 0.30.14

2022-07-10T16:58:00-04:00

Serd 0.30.14 has been released. Serd is a lightweight C library for working with RDF data.

Changes:

Fix memory consumption when reading documents
Switch to Meson build system
Update README and project metadata

Sratom 0.6.10

2022-05-26T23:27:00-04:00

Sratom 0.6.10 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Fix documentation installation directory
Fix potential blank node ID truncation

Sord 0.16.10

2022-05-26T23:22:00-04:00

Sord 0.16.10 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Fix Windows build
Fix potential crash or incorrectness issue with GCC 10 again

Serd 0.30.12

2022-05-26T22:47:00-04:00

Serd 0.30.12 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Fix warnings and build issues with clang 13 and VS 2019
Fix writing long literals with triple quotes
Improve documentation style
Support combining several BSD-style command line flags in serdi
Write statements with invalid URI characters in lax mode

Serd 0.30.10

2021-01-20T15:55:00-05:00

Serd 0.30.10 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add fallback configuration if documentation theme is unavailable
Fix SERD_DISABLE_DEPRECATED
Fix building for older MacOS versions on newer MacOS
Fix documentation installation directory

Sratom 0.6.8

2021-01-07T21:36:00-05:00

Sratom 0.6.8 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Fix potential null pointer dereference

Sord 0.16.8

2021-01-07T21:33:00-05:00

Sord 0.16.8 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Clean up code
Fix potential undefined behavior
Fix potentially incorrect search results
Remove the need for a generated configuration header

Serd 0.30.8

2021-01-07T21:19:00-05:00

Serd 0.30.8 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Deprecate serd_uri_to_path()
Don't install API man pages
Fix potential memory error when serialising URIs
Move headers to an include directory
Refuse to write relative URI references to NTriples
Remove the need for a generated configuration header
Remove use of C character class functions that may use locale
Split up and reorganize unit tests
Use aligned allocation via C11 or Windows API where possible

Sratom 0.6.6

2020-09-27T15:18:00-04:00

Sratom 0.6.6 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Fix various minor warnings and other code quality issues

Sord 0.16.6

2020-09-27T11:08:00-04:00

Sord 0.16.6 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Fix potential crash or incorrectness issues with GCC 10
Fix various minor warnings and other code quality issues

Serd 0.30.6

2020-09-27T10:51:00-04:00

Serd 0.30.6 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Fix various minor warnings and other code quality issues
Improve reader error handling
Make all free functions tolerant of NULL
Parse prefixed names more strictly

Serd 0.30.4

2020-04-26T16:04:00-04:00

Serd 0.30.4 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Fix EOF handling while reading in bulk or from strings
Fix lax handling of string errors
Fix reading from a null-delimited socket

Sord 0.16.4

2019-11-10T21:26:00-05:00

Sord 0.16.4 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Update build system

Sratom 0.6.4

2019-11-10T21:20:00-05:00

Sratom 0.6.4 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Make sratom_free() safe to call on NULL
Various minor code cleanups

Serd 0.30.2

2019-10-20T22:28:00-04:00

Serd 0.30.2 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Fix GCC 4 build
Fix colliding blank nodes when parsing TriG
Fix missing parse error messages
Fix parsing "a" abbreviation without padding whitespace
Fix parsing TriG graphs with several squashed trailing dots
Fix resolving some URIs against base URIs with no trailing slash
Improve build system and CI integration
Improve documentation

Sratom 0.6.2

2018-07-22T18:48:00-04:00

Sratom 0.6.2 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Various minor code cleanups

Sord 0.16.2

2018-07-22T18:38:00-04:00

Sord 0.16.2 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

sord_validate: Add support for maximum cardinality
sord_validate: Fix restriction count reporting
sord_validate: Only report functional / inverse functional property errors if the subjects / objects are not blank
sord_validate: Use consistent error format

Serd 0.30.0

2018-07-22T18:34:00-04:00

Serd 0.30.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add NTriples test suite
Add basic error tolerant parsing for line-based formats
Add serd_free for freeing memory allocated by serd
Add serdi option to write ASCII output
Clarify errors returned by serd_env_expand()
Clean up testing code
Fix building with MSVC
Fix length of base64 encoded nodes
Fix parsing local names that end with escaped dots
Fix reported error when reading statements with only a blank node
Improve relative URI serialisation
Make serdi guess input syntax from extension if unspecified
Make serdi syntax options case-insensitive
Report errors for all failing test cases
Support strict parsing of prefixed names
Upgrade to waf 2.0.9 and autowaf a328983
Use more efficient stack allocation and growth policies

Serd 0.28.0

2017-07-21T06:28:00-04:00

Serd 0.28.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add serd_node_from_substring()
Add support for reading from a user provided callback
Fix hangs when reading corrupt UTF-8
Fix parsing of hex escapes in file URIs (thanks Johannes Mueller)
Fix strict parsing of abolute URI schemes
Gracefully handle applications that write corrupt UTF-8

Serd 0.26.0

2017-01-10T01:10:00-05:00

Serd 0.26.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add support for TriG and NQuads
Add support for braindead SPARQL PREFIX and BASE directives

Sratom 0.6.0

2016-09-20T02:25:00-04:00

Sratom 0.6.0 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Add sratom_set_env() for setting prefixes
Fix padding of constructed vectors (thanks Hanspeter Portner)
Fix warnings when building with ISO C++ compilers
Support round-trip serialisation of relative paths
Support sequences with beat time stamps
Upgrade to waf 1.8.14

Sord 0.16.0

2016-09-20T02:23:00-04:00

Sord 0.16.0 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Add sord_node_get_string_measured() to get both byte and character counts
Fix construction and comparison of URIs with UTF-8 characters
Fix documentation generation
Make sord_node_get_counted return byte count, and clarify documentation
Return error from sord_inserter_write_statement() if a node can not be written (e.g. undefined prefix)
Safely return NULL from sord_iter_get() for end iterators
sord_validate: Do proper numeric comparison for propery bounds checking
sord_validate: Tolerate xsd:decimal literals for double and float properties if literals match pattern

Serd 0.24.0

2016-09-20T02:21:00-04:00

Serd 0.24.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add serd_node_new_relative_uri()
Fix construction and comparison of URIs with UTF-8 characters
Fix documentation generation
Fix potential out of bounds read
Fix unaligned memory access, undefined behaviour which breaks on ARM
Report I/O errors with message and return appropriate status code
Report missing statement separator errors
Update serdi man page

Sord 0.14.0

2015-10-08T19:37:00-04:00

Sord 0.14.0 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Add sord_erase() for erasing statements via an iterator
Fix bugs with stores that contain both graphs and default graph statements
Fix compilation on compilers that do not support -pthread flag
Fix crash caused by multiple deletion of datatype nodes
Fix minor memory leak in sordi
Fix using sordi with stdin
Improve test coverage
Reduce memory usage and increase performance with a better data structure
Show sordi errors in standard format
This release does not break the ABI, but the semantics of iterators has changed: any modification to a model invalidates iterators on that model
Upgrade to waf 1.8.14
sord_validate: More extensive validation, including cardinality, PlainLiteral, and someValuesFrom restrictions.

Serd 0.22.0

2015-10-08T19:34:00-04:00

Serd 0.22.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add serd_reader_set_strict() and -l (lax) option to serdi to tolerate parsing URIs with escaped characters
Add support for Turtle named inline nodes extension
Fix clash resolution when a blank node ID prefix is set
Fix reading statements ending with a blank then dot with no space
Fix serd_reader_read_file() for URIs with escaped characters (spaces)
Fix serializing fractional decimals that would round up
Fix warnings when building with ISO C++ compilers
Remove dependence on fmax() to avoid portability issues
Report errors for invalid IRI characters and missing terminators
Show serdi errors in standard format
Upgrade to waf 1.8.14

Sratom 0.4.6

2014-08-08T22:14:00-04:00

Sratom 0.4.6 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Don't set eg prefix in sratom_to_turtle
Update for latest LV2 Atom Object simplification
Upgrade to waf 1.7.16

Sord 0.12.2

2014-08-08T22:03:00-04:00

Sord 0.12.2 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Fix iteration over an entire graph (* * * graph)
Update to waf 1.7.16
sordmm.hpp: Remove unused members

Serd 0.20.0

2014-08-08T22:02:00-04:00

Serd 0.20.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add -lm to pkg-config libs
Correctly handle posix_memalign failure
Don't write xsd:decimal literals to Turtle bare if they would not be read back with the same type
Fix const-correctness violation for reader input string
Fix possible crash in serd_writer_end_anon() when writing invalid lists
Generate blank names like :b1 and :B2 not :genid1 :docid2
Support new RDF 1.1 Turtle
Update to waf 1.7.14

Sratom 0.4.4

2014-01-04T21:09:00-05:00

Sratom 0.4.4 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Escape invalid characters when writing paths as URIs
Fancier printing without subject and predicate

Sratom 0.4.2

2013-02-18T16:17:00-05:00

Sratom 0.4.2 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Disable timestamps in HTML documentation for reproducible build
Fix memory leaks
Fix serialisation of nested tuples
Update to waf 1.7.9 and autowaf r90 (install docs to versioned directory)
Use new sord API to clean up and shrink code

Sord 0.12.0

2013-02-18T16:05:00-05:00

Sord 0.12.0 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Add sord_get() for easily getting single property values
Update to waf 1.7.9 and autowaf r90 (install docs to versioned directory)
sord_validate: Pass type check when range is xsd:anyURI and value is a URI
sord_validate: Support any subClassOf rdf:Property, not just baked-in ones
sordmm.hpp: Add Node::to_serd_node()
sordmm.hpp: Add convenient constructors for decimal and integer literals
sordmm.hpp: Don't automatically add RDF namespace prefix to world

Serd 0.18.2

2012-12-23T02:32:00-05:00

Serd 0.18.2 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Disable timestamps in HTML documentation for reproducible build
Fix bug that caused "a" abbreviation in non-predicate position
Fix clashing symbol "error" in amalgamation build
Fix crash when resolving against non-standard base URIs
Fix crash when serd_node_new_decimal is called with infinity or NaN
Update to waf 1.7.8 and autowaf r90 (install docs to versioned directory)

Sord 0.10.4

2012-10-14T22:23:00-04:00

Sord 0.10.4 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Disable timestamps in HTML documentation for reproducible build
Fix memory leaks in sord_validate
Implement better data type validation in sord_validate conformant with the XSD and OWL specifications
Install sord_validate man page

How to Define a Datatype in RDF

2012-09-08T16:42:00-04:00

I had to do some digging around to figure out how to define a new Datatype with restrictions in RDF, so I thought it might make a useful post to save someone else the trouble in the future.

RDF datatypes are based on XSD datatypes, which are often used directly. Unfortunately, most implementations simply have the XSD types baked in and do not support or validate new datatype descriptions (though at least sord_validate can). Regardless, it is sometimes necessary to define a datatype with a specific restriction so it can be machine validated. It's a bit tricky to figure out how to do this, since everything is buried in specifications that aren't as triple oriented as they should be. So, here is an example of defining a datatype restricted by regular expression in Turtle, derived from the OWL documentation:

<http://example.org/CSymbol>
    a rdfs:Datatype ;
    rdfs:comment "A symbol in the C programming language" ;
    owl:onDatatype xsd:string ;
    owl:withRestrictions (
        [
            xsd:pattern "[_a-zA-Z][_a-zA-Z0-9]*"
        ]
        ) .

The XSD specification defines several “constraining facets” you can use in this way. See the XSD specification for details, but the most obvious and useful for RDF are: xsd:length, xsd:minLength, xsd:maxLength, xsd:pattern, xsd:maxInclusive, xsd:maxExclusive, xsd:minInclusive, xsd:minExclusive. For example, you can define a numeric type with restricted range like so:

<http://example.org/AnswerishInteger>
    a rdfs:Datatype ;
    rdfs:comment "An integer between 24 and 42 inclusive" ;
    owl:onDatatype xsd:integer ;
    owl:withRestrictions (
        [
            xsd:minInclusive 24
        ] [
            xsd:maxInclusive 42
        ]
    ) .

Defining datatypes in this way and using them as the rdfs:range for properties is a good idea because it describes which values are valid in a machine readable way. This makes it possible for simple generic tools to validate data, ensuring that all literals are valid values for the property they describe.

Sratom 0.4.0

2012-08-23T05:11:00-04:00

Sratom 0.4.0 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Changes:

Correctly read objects with several rdf:type properties
Fix various hyper-strict warnings
Support writing Object Atoms as top level descriptions if subject and predicate are not given.
Upgrade to waf 1.7.2

Sord 0.10.0

2012-08-23T04:19:00-04:00

Sord 0.10.0 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Add error callback to world for custom error reporting
Add option to build utilities as static binaries
Do not require a C++ compiler to build
Fix various hyper-strict warnings
Make all 'zix' symbols private to avoid symbol clashes in static builds
Performance and space (per node) improvements
Remove problematic "Loaded n statements" output from serdi
SSE4.2 accelerated hashing for node interning, where available
Strip down API documentation to a single clean page
Upgrade to waf 1.7.2
sordmm.hpp: Add indices and graphs parameters to Model constructor
sordmm.hpp: Correctly handle Sord::Node self-assignment
sordmm.hpp: Remove overzealous URI scheme assertion

Serd 0.18.0

2012-08-23T04:18:00-04:00

Serd 0.18.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add -e option to serdi to use incremental reading
Add -q option to serdi to suppress all non-data output, e.g. errors
Add error callback to reader and writer for custom error reporting
Add incremental read interface suitable for reading from infinite streams
Add option to build utilities as static binaries
Do not require a C++ compiler to build
Fix various hyper-strict warnings
Report write size correctly when invalid UTF-8 is encountered and a replacement character is written
Reset indent when finishing a write
Strip down API documentation to a single clean page
Support digits at start of local names as per new Turtle grammar
Upgrade to waf 1.7.2

Sratom 0.2.0

2012-04-18T19:14:00-04:00

Sratom 0.2.0 has been released. Sratom is a small library for serialising LV2 atoms to and from RDF, for converting between binary and text or storing in a model. For more information, see http://drobilla.net/software/sratom.

Sord 0.8.0

2012-04-17T22:24:00-04:00

Sord 0.8.0 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Add SordInserter for writing to a model via Serd sink functions.
Add convenient sord_search(), sord_ask(), and sord_count()
Add sord_iter_get_node()
Add sord_new_relative_uri()
Add sord_validate tool for validating data against RDF/OWL schemas
Fix comparison of typed literals
Install man page to DATADIR (e.g. PREFIX/share/man, not PREFIX/man)
Refuse to intern relative URIs in sord_new_uri*()
Support compilation as C++ under MSVC++.
Take advantage of interning in sord_node_equals()
Tolerate serd passing NULL nodes to reader callback (serd 0.6.0)
Use path variables in pkgconfig files

Serd 0.14.0

2012-04-17T22:23:00-04:00

Serd 0.14.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add SerdBulkSink for writing bulk output and corresponding serdi -B option
Add serd_chunk_sink for easy writing to a string
Add serd_file_sink for easy writing to a FILE* stream
Add serd_node_new_blob and serd_base64_decode for handling arbitrary binary data via base64 encoding
Add serd_node_new_file_uri() and serd_file_uri_parse() and implement proper URI to/from path hex escaping, etc.
Add serd_reader_set_default_graph() for reading a file as a named graph
Add serd_strtod(), serd_node_new_decimal(), and serd_node_new_integer() for locale-independent numeric node parsing/serialising
Add serd_uri_serialise_relative() for making URIs relative to a base where possible (by chopping a common prefix and adding dot segments)
Add serd_writer_get_env()
Add serd_writer_set_root_uri() and corresponding -r option to serdi to enable writing URIs with up references (../)
Add serdi -f option to prevent URI qualification
Escape ASCII control characters in output (e.g. fix problems with string literals that start with a backspace)
Handle a quote as the last character of a long string literal in the writer (by escaping it) rather than the reader, to avoid writing Turtle other tools fail to parse
Handle files and strings that start with a UTF-8 Byte Order Mark
Implement pretty-printing for collections
Improve URI resolution to cover most of the abnormal cases from RFC3986
Improve write performance by doing bulk writes for unescaped substrings
Install man page to DATADIR (e.g. PREFIX/share/man, not PREFIX/man)
Make URIs serialised by the writer properly escape characters
Parse collections iteratively in O(1) space
Remove use of multi-byte peek (readahead) and use exactly 1 page for read buffer (instead of 2)
Report read error if both "genid" and "docid" IDs are found in the same document, to prevent silent merging of distinct blank nodes
Report reason for failure to open file in serdi
Resolve dot segments in serd_uri_resolve() instead of at write time
Support Windows file://c:/foo URIs in serd_uri_to_path() on all platforms
Support compilation as C++ under MSVC++
Support file://localhost/foo URIs in serd_uri_to_path()
Tolerate invalid characters in string literals by replacing with the Unicode replacement character
Use path variables in pkgconfig files

Serd 0.5.0

2011-09-29T04:00:00-04:00

Serd 0.5.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Add ability to build static library
Add serd_env_set_prefix_from_strings for convenience
Add serd_strerror
Avoid writing illegal Turtle names as a result of URI qualifying
Fix erroneously equal SERD_ERR_BAD_SYNTAX and SERD_ERR_BAD_ARG
Fix pretty printing of successive blank descriptions, i.e. "] , ["
Gracefully handle NULL reader sinks

Sord 0.5.0

2011-09-29T04:00:00-04:00

Sord 0.5.0 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Add ability to build static library
Add function sord_contains for checking for a triple pattern
Add function sord_write_iter for writing a queried range
Fix Sord::Namespaces::qualify to no longer chop prefixes
Remove glib dependency

Serd 0.4.2

2011-05-25T23:00:00-04:00

Serd 0.4.2 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Changes:

Fix build system Python 3 compatibility
Fix compilation issues on some systems

Sord 0.4.2

2011-05-25T23:00:00-04:00

Sord 0.4.2 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.

Changes:

Bump Serd dependency to 0.4.0
Fix build system Python 3 compatibility
Fix compilation issues on some systems

Serd 0.4.0

2011-05-25T03:00:00-04:00

Serd 0.4.0 has been released. Serd is a lightweight C library for RDF syntax which supports reading and writing Turtle, TriG, NTriples, and NQuads. Serd is suitable for performance-critical or resource-limited applications, such as serialising very large data sets or embedded systems.

Sord 0.4.0

2011-05-25T03:00:00-04:00

Sord 0.4.0 has been released. Sord is a lightweight C library for storing RDF statements in memory. For more information, see http://drobilla.net/software/sord.