Merge branch 'release-0.4.0'

.gitignore

@@ -26,6 +26,6 @@ dist/libschwa.pc
 libtool
 m4/*.m4
 stamp-h1
-/src/lib/schwa/_base.h
 /src/lib/schwa/_config.h*
+/src/lib/schwa/macros.h
 /src/lib/schwa/port.h

CHANGELOG

@@ -1,3 +1,12 @@
+0.4.0
+* Don't distribute _config.h (whoops).
+* Added `SCHWA_LIKELY` and `SCHWA_UNLIKELY` macro definitions to wrap around `__builtin_expect`.
+* Made `--help` more helpful by adding a `--short-help` flag on all core tools (#14).
+* Made `dr-count` behave more like `wc` when multiple input files are provided (#15).
+* Corrected argument parsing for optional arguments (#17).
+* Upgraded the Unicode database data from 6.3.0 to 7.0.0.
+* Upgraded docrep reader and writer to use wire protocol version 3 (utilising the newly-added distinction between bytes and UTF-8 strings within MessagePack).
+
 0.3.0
 * Allowed `dr-count` and `dr-grep` to skip incomplete documents (#6).
 * Corrected bug lazy docrep reading to ensure that it reads the desired number of bytes (#5).

Makefile.am

@@ -8,7 +8,7 @@ SUBDIRS = src/third-party src/lib src/apps
 pkgconfigdir = $(libdir)/pkgconfig
 pkgconfig_DATA = dist/libschwa.pc
 
-EXTRA_DIST = LICENCE doc src/ragel src/third-party/cityhash src/third-party/re2
+EXTRA_DIST = CHANGELOG LICENCE doc src/ragel src/third-party/cityhash src/third-party/re2
 dist-hook: clean-docs
 
 
@@ -27,5 +27,5 @@ clean-docs:
 
 wc:
 	for d in src/lib src/apps; do \
-		find $${d} \( -name "*.cc" -or -name "*.h" \) | egrep -v '_(gen|test)\.cc' | xargs wc -l; \
+		find $${d} \( -name "*.cc" -or -name "*.h" \) | egrep -v '_(gen|test)\.cc' | grep -vF /third-party/ | xargs wc -l; \
 	done

README

@@ -1 +1,3 @@
+Schwa Lab core NLP tools, including the C++ implementation of docrep and the docrep command-line tools.
+
 Documentation can be found on the GitHub wiki page: https://github.com/schwa-lab/libschwa/wiki

configure.ac

@@ -3,7 +3,7 @@ AC_INIT([libschwa], m4_normalize(m4_include([m4/version.m4])), [[email protected]
 : ${CXXFLAGS=""}
 
 m4_include([m4/version-set.m4])
-AC_SUBST([libschwa_la_version_info], [3:0:0])
+AC_SUBST([libschwa_la_version_info], [4:0:0])
 
 AC_CONFIG_AUX_DIR([build-aux])
 AC_CONFIG_MACRO_DIR([m4])
@@ -156,6 +156,11 @@ AC_COMPILE_IFELSE(
 )
 if test x$libschwa_have_builtin_expect = xyes ; then
   AC_DEFINE([HAVE_BUILTIN_EXPECT], [1], [Define to 1 if the compiler supports __builtin_expect.])
+  AC_SUBST([SCHWA_LIKELY], ['(__builtin_expect(!!(x), 1))'])
+  AC_SUBST([SCHWA_UNLIKELY], ['(__builtin_expect(!!(x), 0))'])
+else
+  AC_SUBST([SCHWA_LIKELY], ['(x)'])
+  AC_SUBST([SCHWA_UNLIKELY], ['(x)'])
 fi
 
 dnl Work out how to inline the "host to big endian" functions for various based on what headers we found.
@@ -219,7 +224,7 @@ AC_CONFIG_FILES([
   src/apps/dr-worker-example/Makefile
   src/apps/schwa-tokenizer/Makefile
   src/lib/Makefile
-  src/lib/schwa/_base.h:src/lib/schwa/_base.h.in
+  src/lib/schwa/macros.h:src/lib/schwa/macros.h.in
   src/lib/schwa/port.h:src/lib/schwa/port.h.in
   src/third-party/Makefile
   dist/libschwa.pc:dist/libschwa.pc.in

dist/libschwa.rb

@@ -2,8 +2,8 @@
 
 class Libschwa < Formula
   homepage "https://github.com/schwa-lab/libschwa"
-  url "https://github.com/schwa-lab/libschwa/releases/download/0.2.2/libschwa-0.2.2.tar.gz"
-  sha1 "41d3e1a2dd3686da6ee0acb302e88eb84ca937bb"
+  url "https://github.com/schwa-lab/libschwa/releases/download/0.3.0/libschwa-0.3.0.tar.gz"
+  sha1 "9b44fb99547b3e8dd1a8e1e7bf24615f3f437bd6"
 
   depends_on "pkg-config"
   depends_on "zeromq" => :optional

m4/version-set.m4

@@ -1,13 +1,13 @@
 dnl Product version for header files. Maintained by dist/bump-version.sh
 VERSION_MAJOR=0
-VERSION_MINOR=3
-VERSION_PATCH=1
-VERSION_STRING='"0.3.1 (2014-07-06)"'
+VERSION_MINOR=4
+VERSION_PATCH=0
+VERSION_STRING='"0.4.0 (2014-09-05)"'
 
 AC_SUBST(VERSION_MAJOR)
 AC_SUBST(VERSION_MINOR)
 AC_SUBST(VERSION_PATCH)
 AC_SUBST(VERSION_STRING)
 
-VERSION_NOPATCH=0.3
+VERSION_NOPATCH=0.4
 AC_SUBST(VERSION_NOPATCH)

m4/version.m4

@@ -1,2 +1,2 @@
 dnl Product version for AC_INIT. Maintained by dist/bump-version.sh
-0.3.1
+0.4.0

src/apps/dr-count/main.cc

@@ -14,16 +14,14 @@ namespace cf = schwa::config;
 namespace dr = schwa::dr;
 namespace io = schwa::io;
 
-using Formatting = schwa::dr_count::Processor::Formatting;
-
 
 namespace schwa {
 namespace dr_count {
 
 static void
-main(const std::vector<std::string> &input_paths, std::ostream &out, bool all_stores, const std::string &store, bool count_bytes, bool cumulative, bool per_doc, Formatting formatting, const std::string &doc_id) {
+main(const std::vector<std::string> &input_paths, std::ostream &out, bool all_stores, const std::string &store, bool count_bytes, bool cumulative, int every, Formatting formatting, const std::string &doc_id, bool no_header, bool no_footer, bool no_ndocs) {
   // Construct the document processor.
-  Processor processor(out, all_stores, store, count_bytes, cumulative, per_doc, formatting, doc_id);
+  Processor processor(out, all_stores, store, count_bytes, cumulative, every, formatting, doc_id, no_header, no_footer, no_ndocs);
 
   dr::FauxDoc doc;
   dr::FauxDoc::Schema schema;
@@ -34,6 +32,12 @@ main(const std::vector<std::string> &input_paths, std::ostream &out, bool all_st
     io::InputStream in(input_path);
     dr::Reader reader(*in, schema);
 
+    // If there is more than one file to process, output the filename and reset the processors state.
+    if (input_paths.size() > 1) {
+      out << input_path << std::endl;
+      processor.reset();
+    }
+
     // Read the documents off the input stream.
     try {
       while (reader >> doc)
@@ -42,9 +46,9 @@ main(const std::vector<std::string> &input_paths, std::ostream &out, bool all_st
     catch (dr::ReaderException &) {
       LOG(WARNING) << "Failed to read document from '" << in.path() << "'" << std::endl;
     }
-  }
 
-  processor.finalise();
+    processor.finalise();
+  }
 }
 
 }  // namespace dr_count
@@ -53,17 +57,22 @@ main(const std::vector<std::string> &input_paths, std::ostream &out, bool all_st
 
 int
 main(int argc, char **argv) {
+  using Formatting = schwa::dr_count::Formatting;
+
   // Construct an option parser.
   cf::Main cfg("dr-count", "Count the number of documents or annotations in stores on a docrep stream.");
   cf::Op<std::string> input_path(cfg, "input", 'i', "The input path", io::STDIN_STRING);
   cf::Op<std::string> output_path(cfg, "output", 'o', "The output path", io::STDOUT_STRING);
   cf::Op<bool> all_stores(cfg, "all", 'a', "Count docs and elements in all stores found on the first doc", false);
   cf::Op<std::string> store(cfg, "store", 's', "Count docs and elements in the provided store only", cf::Flags::OPTIONAL);
-  cf::Op<bool> per_doc(cfg, "per-doc", 'e', "Show counts per doc instead of for the stream", false);
+  cf::Op<uint32_t> every(cfg, "every", 'e', "Show counts per doc instead of for the stream", 1, cf::Flags::OPTIONAL);
   cf::Op<bool> count_bytes(cfg, "bytes", 'b', "Count bytes instead of number of instances", false);
   cf::Op<bool> cumulative(cfg, "cumulative", 'c', "Show cumulative counts per doc", false);
   cf::OpChoices<std::string> format(cfg, "format", 'f', "How to format the output data", {"aligned", "tabs"}, "aligned");
   cf::Op<std::string> doc_id(cfg, "doc-id", 'd', "Output this expression before each document instead when outputting per-document counts", cf::Flags::OPTIONAL);
+  cf::Op<bool> no_header(cfg, "no-header", 'H', "Hide the column headings row", false);
+  cf::Op<bool> no_footer(cfg, "no-footer", 'F', "Hide the cumulative total footer row", false);
+  cf::Op<bool> no_ndocs(cfg, "no-ndocs", 'N', "Hide the ndocs column", false);
 
   cfg.allow_unclaimed_args("[input-path...]");
 
@@ -89,7 +98,7 @@ main(int argc, char **argv) {
     io::OutputStream out(output_path());
 
     // Dispatch to main function.
-    schwa::dr_count::main(input_paths, *out, all_stores(), store(), count_bytes(), cumulative(), per_doc(), formatting, doc_id());
+    schwa::dr_count::main(input_paths, *out, all_stores(), store(), count_bytes(), cumulative(), every.was_mentioned() ? every() : -1, formatting, doc_id(), no_header(), no_footer(), no_ndocs());
   })
   return 0;
 }