mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2024-11-24 02:59:07 +08:00
Code Issues Actions 10 Packages Projects Releases Wiki Activity

Fix and enable dawg_test

Browse Source 
Signed-off-by: Stefan Weil <sw@weilnetz.de>
This commit is contained in:
Stefan Weil 2019-01-21 11:47:10 +01:00
parent aec992ebf8
commit 05cdbc7c9c
2 changed files with 25 additions and 38 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
unittest/Makefile.am
View File

@ -1,6 +1,3 @@
# Absolute path of directory 'src'.
TESS_SRC_DIR=$(shell cd $(top_srcdir) && pwd)/src
# Absolute path of directory 'langdata'.
LANGDATA_DIR=$(shell cd $(top_srcdir) && cd .. && pwd)/langdata_lstm
@ -15,7 +12,7 @@ TESTING_DIR=$(shell cd $(top_srcdir) && pwd)/test/testing
# (using submodule test).
TESTDATA_DIR=$(shell cd $(top_srcdir) && pwd)/test/testdata
AM_CPPFLAGS += -DTESS_SRC_DIR="\"$(TESS_SRC_DIR)\""
AM_CPPFLAGS += -DTESSBIN_DIR="\"$(abs_top_builddir)\""
AM_CPPFLAGS += -DLANGDATA_DIR="\"$(LANGDATA_DIR)\""
AM_CPPFLAGS += -DTESSDATA_DIR="\"$(TESSDATA_DIR)\""
AM_CPPFLAGS += -DTESTING_DIR="\"$(TESTING_DIR)\""
@ -107,6 +104,7 @@ check_PROGRAMS = \
bitvector_test \
cleanapi_test \
colpartition_test \
dawg_test \
denorm_test \
fileio_test \
heap_test \

57
unittest/dawg_test.cc
View File

@ -1,4 +1,6 @@
#include <cstdlib> // for system
#include <fstream> // for ifstream
#include <set>
#include <string>
#include <vector>
@ -8,55 +10,42 @@
#include "trie.h"
#include "include_gunit.h"
#include "base/filelinereader.h"
#include "util/process/subprocess.h"
namespace {
void RemoveTrailingLineTerminators(char* line) {
char* end = line + strlen(line) - 1;
while (end >= line && ('\n' == *end || '\r' == *end)) {
*end-- = 0;
}
}
void AddLineToSet(std::set<std::string>* words, char* line) {
RemoveTrailingLineTerminators(line);
words->insert(line);
}
// Test some basic functionality dealing with Dawgs (compressed dictionaries,
// aka Directed Acyclic Word Graphs).
class DawgTest : public testing::Test {
protected:
void LoadWordlist(const std::string& filename, std::set<std::string>* words) const {
FileLineReader::Options options;
options.set_comment_char(0);
FileLineReader flr(filename.c_str(), options);
flr.set_line_callback(NewPermanentCallback(AddLineToSet, words));
flr.Reload();
std::ifstream file(filename);
if (file.is_open()) {
std::string line;
while (getline(file, line)) {
// Remove trailing line terminators from line.
while (!line.empty() && (line.back() == '\n' || line.back() == '\r')) {
line.resize(line.size() - 1);
}
// Add line to set.
words->insert(line.c_str());
}
file.close();
}
}
std::string TestDataNameToPath(const std::string& name) const {
return file::JoinPath(TESTDATA_DIR, "/" + name);
return file::JoinPath(TESTDATA_DIR, name);
}
std::string TessBinaryPath(const std::string& binary_name) const {
return file::JoinPath(TESS_SRC_DIR,
std::string TessBinaryPath(const std::string& name) const {
return file::JoinPath(TESSBIN_DIR, "src/training/" + name);
}
std::string OutputNameToPath(const std::string& name) const {
return file::JoinPath(FLAGS_test_tmpdir, name);
}
int RunCommand(const std::string& program, const std::string& arg1, const std::string& arg2,
const std::string& arg3) const {
SubProcess p;
std::vector<std::string> argv;
argv.push_back(program);
argv.push_back(arg1);
argv.push_back(arg2);
argv.push_back(arg3);
p.SetProgram(TessBinaryPath(program), argv);
p.Start();
p.Wait();
return p.exit_code();
int RunCommand(const std::string& program, const std::string& arg1,
const std::string& arg2, const std::string& arg3) const {
std::string cmdline =
TessBinaryPath(program) + " " + arg1 + " " + arg2 + " " + arg3;
return system(cmdline.c_str());
}
// Test that we are able to convert a wordlist file (one "word" per line) to
// a dawg (a compressed format) and then extract the original wordlist back