diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..8be7a7b --- /dev/null +++ b/.gitignore @@ -0,0 +1 @@ +SOURCES/libkkc-data-0.2.7.tar.xz diff --git a/.libkkc-data.metadata b/.libkkc-data.metadata new file mode 100644 index 0000000..df3ddfd --- /dev/null +++ b/.libkkc-data.metadata @@ -0,0 +1 @@ +6e5d833744f8311908f12168dcf697633efc2795 SOURCES/libkkc-data-0.2.7.tar.xz diff --git a/SOURCES/libkkc-data-HEAD.patch b/SOURCES/libkkc-data-HEAD.patch new file mode 100644 index 0000000..af52f9a --- /dev/null +++ b/SOURCES/libkkc-data-HEAD.patch @@ -0,0 +1,128 @@ +diff --git a/tools/genfilter.py b/tools/genfilter.py +index 5ffab32..0c5f75a 100644 +--- a/tools/genfilter.py ++++ b/tools/genfilter.py +@@ -84,24 +84,24 @@ class FilterGenerator(object): + + def generate(self): + size = os.fstat(self.infile.fileno()).st_size +- n = size / self.record_size ++ n = size // self.record_size + m = int(math.ceil(-n*math.log10(ERROR_RATE) / + math.pow(math.log10(2), 2))) +- m = (m/8 + 1)*8 ++ m = (m//8 + 1)*8 + inmem = mmap.mmap(self.infile.fileno(), + size, + access=mmap.ACCESS_READ) +- outmem = bytearray(m/8) +- for i in xrange(0, n): ++ outmem = bytearray(m//8) ++ for i in range(0, n): + offset = i*self.record_size + b0, b1 = struct.unpack("=LL", inmem[offset:offset+8]) +- for k in xrange(0, 4): ++ for k in range(0, 4): + h = murmur_hash3_32(b0, b1, k) + h = int(h * (m / float(0xFFFFFFFF))) +- outmem[h/8] |= (1 << (h%8)) ++ outmem[h//8] |= (1 << (h%8)) + inmem.close() +- # Convert bytearray to str, for Python 2.6 compatibility. +- self.outfile.write(str(outmem)) ++ # Convert bytearray to bytes, for Python 3 compatibility. ++ self.outfile.write(bytes(outmem)) + + if __name__ == '__main__': + import sys +@@ -110,7 +110,7 @@ if __name__ == '__main__': + parser = argparse.ArgumentParser(description='filter') + parser.add_argument('infile', type=argparse.FileType('r'), + help='input file') +- parser.add_argument('outfile', type=argparse.FileType('w'), ++ parser.add_argument('outfile', type=argparse.FileType('wb'), + help='output file') + parser.add_argument('record_size', type=int, + help='record size') +diff --git a/tools/sortlm.py b/tools/sortlm.py +index a0dd8fe..40f0837 100644 +--- a/tools/sortlm.py ++++ b/tools/sortlm.py +@@ -40,10 +40,10 @@ class SortedGenerator(object): + self.__min_cost = 0.0 + + def read(self): +- print "reading N-grams" ++ print("reading N-grams") + self.__read_tries() + self.__read_ngrams() +- print "min cost = %lf" % self.__min_cost ++ print("min cost = %lf" % self.__min_cost) + + def __read_tries(self): + while True: +@@ -58,7 +58,7 @@ class SortedGenerator(object): + line = self.__infile.readline() + if line == "": + break +- line = line.strip() ++ line = line.strip('\n') + if line == "": + break + match = self.__ngram_line_regex.match(line) +@@ -89,7 +89,7 @@ class SortedGenerator(object): + line = self.__infile.readline() + if line == "": + break +- line = line.strip() ++ line = line.strip('\n') + if line == "": + break + match = self.__ngram_line_regex.match(line) +@@ -125,14 +125,11 @@ class SortedGenerator(object): + def quantize(cost, min_cost): + return max(0, min(65535, int(cost * 65535 / min_cost))) + +- def cmp_header(a, b): +- return cmp(a[0], b[0]) +- +- print "writing 1-gram file" ++ print("writing 1-gram file") + unigram_offsets = {} + unigram_file = open("%s.1gram" % self.__output_prefix, "wb") + offset = 0 +- for ids, value in sorted(self.__ngram_entries[0].iteritems()): ++ for ids, value in sorted(self.__ngram_entries[0].items()): + unigram_offsets[ids[0]] = offset + s = struct.pack("=HHH", + quantize(value[0], self.__min_cost), +@@ -143,13 +140,13 @@ class SortedGenerator(object): + offset += 1 + unigram_file.close() + +- print "writing 2-gram file" ++ print("writing 2-gram file") + bigram_offsets = {} + bigram_file = open("%s.2gram" % self.__output_prefix, "wb") + keys = self.__ngram_entries[1].keys() + items = [(struct.pack("=LL", ids[1], unigram_offsets[ids[0]]), ids) for ids in keys] + offset = 0 +- for header, ids in sorted(items, cmp=cmp_header): ++ for header, ids in sorted(items, key=lambda x: x[0]): + value = self.__ngram_entries[1][ids] + bigram_offsets[ids] = offset + s = struct.pack("=HH", +@@ -160,11 +157,11 @@ class SortedGenerator(object): + bigram_file.close() + + if len(self.__ngram_entries[2]) > 0: +- print "writing 3-gram file" ++ print("writing 3-gram file") + trigram_file = open("%s.3gram" % self.__output_prefix, "wb") + keys = self.__ngram_entries[2].keys() + items = [(struct.pack("=LL", ids[2], bigram_offsets[(ids[0], ids[1])]), ids) for ids in keys] +- for header, ids in sorted(items, cmp=cmp_header): ++ for header, ids in sorted(items, key=lambda x: x[0]): + value = self.__ngram_entries[2][ids] + s = struct.pack("=H", + quantize(value[0], self.__min_cost)) diff --git a/SPECS/libkkc-data.spec b/SPECS/libkkc-data.spec new file mode 100644 index 0000000..0665aca --- /dev/null +++ b/SPECS/libkkc-data.spec @@ -0,0 +1,82 @@ +%global debug_package %{nil} + +Name: libkkc-data +Version: 0.2.7 +Release: 11%{?dist} +Epoch: 1 +Summary: Language model data for libkkc + +License: GPLv3+ +URL: https://bitbucket.org/libkkc +Source0: https://bitbucket.org/libkkc/libkkc-data/downloads/%{name}-%{version}.tar.xz + +# Upstream commit https://github.com/ueno/libkkc/commit/ba1c1bd3eb86d887fc3689c3142732658071b5f7 +Patch0: libkkc-data-HEAD.patch + +BuildRequires: python3-devel +BuildRequires: python3-marisa + +%description +The %{name} package contains the language model data that libkkc uses +at run time. + + +%prep +%setup -q +%patch0 -p1 -b .HEAD + + +%build +%configure --disable-static PYTHON=python3 +make %{?_smp_mflags} + + +%install +%make_install INSTALL="install -p" + + +%files +%doc COPYING +%{_libdir}/libkkc + + +%changelog +* Tue Aug 14 2018 Parag Nemade - 1:0.2.7-11 +- Resolves:rh#1615534: libkkc-data FTBFS for missing BuildRequires +- moved this package to use python3 + +* Wed Feb 07 2018 Fedora Release Engineering - 1:0.2.7-10 +- Rebuilt for https://fedoraproject.org/wiki/Fedora_28_Mass_Rebuild + +* Thu Aug 03 2017 Fedora Release Engineering - 1:0.2.7-9 +- Rebuilt for https://fedoraproject.org/wiki/Fedora_27_Binutils_Mass_Rebuild + +* Wed Jul 26 2017 Fedora Release Engineering - 1:0.2.7-8 +- Rebuilt for https://fedoraproject.org/wiki/Fedora_27_Mass_Rebuild + +* Fri Feb 10 2017 Fedora Release Engineering - 1:0.2.7-7 +- Rebuilt for https://fedoraproject.org/wiki/Fedora_26_Mass_Rebuild + +* Thu Feb 04 2016 Fedora Release Engineering - 1:0.2.7-6 +- Rebuilt for https://fedoraproject.org/wiki/Fedora_24_Mass_Rebuild + +* Wed Jun 17 2015 Fedora Release Engineering - 1:0.2.7-5 +- Rebuilt for https://fedoraproject.org/wiki/Fedora_23_Mass_Rebuild + +* Sun Aug 17 2014 Fedora Release Engineering - 1:0.2.7-4 +- Rebuilt for https://fedoraproject.org/wiki/Fedora_21_22_Mass_Rebuild + +* Sat Jun 07 2014 Fedora Release Engineering - 1:0.2.7-3 +- Rebuilt for https://fedoraproject.org/wiki/Fedora_21_Mass_Rebuild + +* Fri Sep 20 2013 Daiki Ueno - 1:0.2.7-2 +- bump release to avoid NVR conflict + +* Fri Sep 20 2013 Daiki Ueno - 1:0.2.7-1 +- add COPYING to %%doc +- disable debuginfo +- add Epoch to avoid conflict with the libkkc package + +* Tue Sep 17 2013 Daiki Ueno - 0.2.7-1 +- initial packaging for Fedora, splitting from libkkc +