unittest: Remove dependency on absl::StripAsciiWhitespace()

This removes the last dependency on Abseil, so that submodule
is now removed completely.
Signed-off-by: NStefan Weil <sw@weilnetz.de>

.github/workflows/unittest-macos.yml

@@ -25,7 +25,7 @@ jobs:
       run: |
            brew install autoconf automake libarchive
            brew install leptonica cairo pango
-           brew install cabextract abseil
+           brew install cabextract
 
     - name: Setup
       run: |

.gitmodules

-[submodule "abseil"]
-	path = abseil
-	url = https://github.com/abseil/abseil-cpp.git
 [submodule "googletest"]
 	path = googletest
 	url = https://github.com/google/googletest.git

CONTRIBUTING.md

@@ -69,7 +69,7 @@ your question has been asked (and has been answered) many times before...
 
 You should always make sure your changes build and run successfully.
 
-For that, your clone needs to have all submodules (`abseil`, `googletest`, `test`) included. To do so, either specify `--recurse-submodules` during the initial clone, or run `git submodule update --init --recursive NAME` for each `NAME` later. If `configure` already created those directories (blocking the clone), remove them first (or `make distclean`), then clone and reconfigure.
+For that, your clone needs to have all submodules (`googletest`, `test`) included. To do so, either specify `--recurse-submodules` during the initial clone, or run `git submodule update --init --recursive NAME` for each `NAME` later. If `configure` already created those directories (blocking the clone), remove them first (or `make distclean`), then clone and reconfigure.
 
 Have a look at [the README](./README.md) and [testing README](./test/testing/README.md) and the [documentation](https://tesseract-ocr.github.io/tessdoc/Compiling-%E2%80%93-GitInstallation.html#unit-test-builds) on installation.
 

Makefile.am

@@ -1156,7 +1156,6 @@ unittest_CPPFLAGS += $(pangocairo_CFLAGS)
 endif # ENABLE_TRAINING
 unittest_CPPFLAGS += -I$(top_srcdir)/src/viewer
 unittest_CPPFLAGS += -I$(top_srcdir)/src/wordrec
-unittest_CPPFLAGS += -I$(top_srcdir)/abseil
 if TENSORFLOW
 unittest_CPPFLAGS += -DINCLUDE_TENSORFLOW
 unittest_CPPFLAGS += -I$(top_srcdir)/unittest
@@ -1170,31 +1169,6 @@ libgtest_la_CPPFLAGS = -I$(top_srcdir)/googletest/googletest/include -I$(top_src
 libgtest_main_la_SOURCES = googletest/googletest/src/gtest_main.cc
 libgtest_main_la_CPPFLAGS = $(libgtest_la_CPPFLAGS)
 
-# Build Abseil (needed for some unit tests).
-check_LTLIBRARIES += libabseil.la
-libabseil_la_SOURCES =
-libabseil_la_SOURCES += abseil/absl/base/internal/raw_logging.cc
-libabseil_la_SOURCES += abseil/absl/base/internal/spinlock.cc
-libabseil_la_SOURCES += abseil/absl/base/internal/spinlock_wait.cc
-libabseil_la_SOURCES += abseil/absl/base/internal/sysinfo.cc
-libabseil_la_SOURCES += abseil/absl/base/internal/throw_delegate.cc
-libabseil_la_SOURCES += abseil/absl/numeric/int128.cc
-libabseil_la_SOURCES += abseil/absl/strings/ascii.cc
-libabseil_la_SOURCES += abseil/absl/strings/charconv.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/charconv_bigint.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/charconv_parse.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/memutil.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/str_format/arg.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/str_format/bind.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/str_format/extension.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/str_format/float_conversion.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/str_format/output.cc
-libabseil_la_SOURCES += abseil/absl/strings/internal/str_format/parser.cc
-libabseil_la_SOURCES += abseil/absl/strings/match.cc
-libabseil_la_SOURCES += abseil/absl/strings/numbers.cc
-libabseil_la_SOURCES += abseil/absl/strings/string_view.cc
-libabseil_la_CPPFLAGS = -I$(top_srcdir)/abseil
-
 GMOCK_INCLUDES = -I$(top_srcdir)/googletest/googlemock/include \
                  -I$(top_srcdir)/googletest/googlemock \
                  -I$(top_srcdir)/googletest/googletest/include \
@@ -1208,7 +1182,6 @@ libgmock_main_la_CPPFLAGS = $(GMOCK_INCLUDES) \
                             -pthread
 
 # Build unittests
-ABSEIL_LIBS = libabseil.la
 GTEST_LIBS =  libgtest.la libgtest_main.la -lpthread
 GMOCK_LIBS =  libgmock.la libgmock_main.la
 TESS_LIBS = $(GTEST_LIBS)
@@ -1330,12 +1303,11 @@ endif # !DISABLED_LEGACY_ENGINE
 
 baseapi_test_SOURCES = unittest/baseapi_test.cc
 baseapi_test_CPPFLAGS = $(unittest_CPPFLAGS)
-baseapi_test_LDADD = $(ABSEIL_LIBS) $(TRAINING_LIBS) $(LEPTONICA_LIBS)
+baseapi_test_LDADD = $(TRAINING_LIBS) $(LEPTONICA_LIBS)
 
 baseapi_thread_test_SOURCES = unittest/baseapi_thread_test.cc
 baseapi_thread_test_CPPFLAGS = $(unittest_CPPFLAGS)
-baseapi_thread_test_LDADD = $(ABSEIL_LIBS)
-baseapi_thread_test_LDADD += $(TESS_LIBS) $(LEPTONICA_LIBS)
+baseapi_thread_test_LDADD = $(TESS_LIBS) $(LEPTONICA_LIBS)
 
 if !DISABLED_LEGACY_ENGINE
 bitvector_test_SOURCES = unittest/bitvector_test.cc
@@ -1432,19 +1404,19 @@ loadlang_test_LDADD = $(TESS_LIBS) $(LEPTONICA_LIBS)
 
 lstm_recode_test_SOURCES = unittest/lstm_recode_test.cc
 lstm_recode_test_CPPFLAGS = $(unittest_CPPFLAGS)
-lstm_recode_test_LDADD = $(ABSEIL_LIBS) $(TRAINING_LIBS)
+lstm_recode_test_LDADD = $(TRAINING_LIBS)
 
 lstm_squashed_test_SOURCES = unittest/lstm_squashed_test.cc
 lstm_squashed_test_CPPFLAGS = $(unittest_CPPFLAGS)
-lstm_squashed_test_LDADD = $(ABSEIL_LIBS) $(TRAINING_LIBS)
+lstm_squashed_test_LDADD = $(TRAINING_LIBS)
 
 lstm_test_SOURCES = unittest/lstm_test.cc
 lstm_test_CPPFLAGS = $(unittest_CPPFLAGS)
-lstm_test_LDADD = $(ABSEIL_LIBS) $(TRAINING_LIBS)
+lstm_test_LDADD = $(TRAINING_LIBS)
 
 lstmtrainer_test_SOURCES = unittest/lstmtrainer_test.cc
 lstmtrainer_test_CPPFLAGS = $(unittest_CPPFLAGS)
-lstmtrainer_test_LDADD = $(ABSEIL_LIBS) $(TRAINING_LIBS) $(LEPTONICA_LIBS)
+lstmtrainer_test_LDADD = $(TRAINING_LIBS) $(LEPTONICA_LIBS)
 
 if !DISABLED_LEGACY_ENGINE
 mastertrainer_test_SOURCES = unittest/mastertrainer_test.cc
@@ -1489,7 +1461,7 @@ pango_font_info_test_SOURCES += unittest/util/utf8/unicodetext.cc
 pango_font_info_test_SOURCES += unittest/util/utf8/unilib.cc
 endif # TENSORFLOW
 pango_font_info_test_CPPFLAGS = $(unittest_CPPFLAGS)
-pango_font_info_test_LDADD = $(ABSEIL_LIBS) $(TRAINING_LIBS) $(LEPTONICA_LIBS)
+pango_font_info_test_LDADD = $(TRAINING_LIBS) $(LEPTONICA_LIBS)
 pango_font_info_test_LDADD += $(ICU_I18N_LIBS)
 pango_font_info_test_LDADD += $(pangocairo_LIBS)
 pango_font_info_test_LDADD += $(pangoft2_LIBS)
@@ -1511,7 +1483,7 @@ progress_test_LDADD = $(GTEST_LIBS) $(GMOCK_LIBS) $(TESS_LIBS) $(LEPTONICA_LIBS)
 
 qrsequence_test_SOURCES = unittest/qrsequence_test.cc
 qrsequence_test_CPPFLAGS = $(unittest_CPPFLAGS)
-qrsequence_test_LDADD = $(ABSEIL_LIBS) $(TESS_LIBS)
+qrsequence_test_LDADD = $(TESS_LIBS)
 
 recodebeam_test_SOURCES = unittest/recodebeam_test.cc
 recodebeam_test_CPPFLAGS = $(unittest_CPPFLAGS)

abseil @ e1d388e7

-Subproject commit e1d388e7e74803050423d035e4374131b9b57919

sw.cpp

@@ -222,7 +222,6 @@ void build(Solution &s)
             t += pango_training;
             t += "org.sw.demo.google.googletest.gmock.main"_dep;
             t += "org.sw.demo.google.googletest.gtest.main"_dep;
-            t += "org.sw.demo.google.abseil"_dep;
 
             if (t.getCompilerType() == CompilerType::MSVC)
                 t.CompileOptions.push_back("-utf-8");

unittest/README.md

@@ -60,7 +60,7 @@
 │   └── script
 │       └── Latin.traineddata
 └── tesseract
-    ├── abseil
+    ├── googletest
     ...
     ├── test
     ├── unittest

unittest/baseapi_test.cc

@@ -19,7 +19,6 @@
 #include <tesseract/baseapi.h>
 
 #include <allheaders.h>
-#include "absl/strings/ascii.h"
 #include "gmock/gmock-matchers.h"
 
 #include <memory>
@@ -48,7 +47,7 @@ std::string GetCleanedTextResult(tesseract::TessBaseAPI *tess, Image pix) {
   char *result = tess->GetUTF8Text();
   std::string ocr_result = result;
   delete[] result;
-  absl::StripAsciiWhitespace(&ocr_result);
+  trim(ocr_result);
   return ocr_result;
 }
 
@@ -80,7 +79,7 @@ TEST_F(TesseractTest, BasicTesseractTest) {
     ocr_text = GetCleanedTextResult(&api, src_pix);
     CHECK_OK(
         file::GetContents(TestDataNameToPath("phototest.gold.txt"), &truth_text, file::Defaults()));
-    absl::StripAsciiWhitespace(&truth_text);
+    trim(truth_text);
     EXPECT_STREQ(truth_text.c_str(), ocr_text.c_str());
     src_pix.destroy();
   } else {
@@ -201,7 +200,7 @@ TEST_F(TesseractTest, AdaptToWordStrTest) {
     Image src_pix = pixRead(TestDataNameToPath(kTestPages[i]).c_str());
     CHECK(src_pix);
     ocr_text = GetCleanedTextResult(&api, src_pix);
-    absl::StripAsciiWhitespace(&truth_text);
+    trim(truth_text);
     EXPECT_STREQ(kTestText[i], ocr_text.c_str());
     src_pix.destroy();
   }
@@ -223,7 +222,7 @@ TEST_F(TesseractTest, BasicLSTMTest) {
   ocr_text = GetCleanedTextResult(&api, src_pix);
   CHECK_OK(
       file::GetContents(TestDataNameToPath("phototest.gold.txt"), &truth_text, file::Defaults()));
-  absl::StripAsciiWhitespace(&truth_text);
+  trim(truth_text);
   EXPECT_STREQ(truth_text.c_str(), ocr_text.c_str());
   src_pix.destroy();
 }

unittest/baseapi_thread_test.cc

@@ -28,7 +28,6 @@
 #endif
 #include <allheaders.h>
 #include <tesseract/baseapi.h>
-#include "absl/strings/ascii.h" // for absl::StripAsciiWhitespace
 #include "commandlineflags.h"
 #include "include_gunit.h"
 #include "log.h"
@@ -148,12 +147,12 @@ static void InitTessInstance(TessBaseAPI *tess, const std::string &lang) {
   EXPECT_EQ(0, tess->Init(TESSDATA_DIR, lang.c_str()));
 }
 
-static void GetCleanedText(TessBaseAPI *tess, Image pix, std::string *ocr_text) {
+static void GetCleanedText(TessBaseAPI *tess, Image pix, std::string &ocr_text) {
   tess->SetImage(pix);
   char *result = tess->GetUTF8Text();
-  *ocr_text = result;
+  ocr_text = result;
   delete[] result;
-  absl::StripAsciiWhitespace(ocr_text);
+  trim(ocr_text);
 }
 
 static void VerifyTextResult(TessBaseAPI *tess, Image pix, const std::string &lang,
@@ -166,7 +165,7 @@ static void VerifyTextResult(TessBaseAPI *tess, Image pix, const std::string &la
     InitTessInstance(tess_local, lang);
   }
   std::string ocr_text;
-  GetCleanedText(tess_local, pix, &ocr_text);
+  GetCleanedText(tess_local, pix, ocr_text);
   EXPECT_STREQ(expected_text.c_str(), ocr_text.c_str());
   if (tess_local != tess) {
     delete tess_local;
@@ -180,7 +179,7 @@ TEST_F(BaseapiThreadTest, TestBasicSanity) {
     TessBaseAPI tess;
     InitTessInstance(&tess, langs_[i]);
     std::string ocr_text;
-    GetCleanedText(&tess, pix_[i], &ocr_text);
+    GetCleanedText(&tess, pix_[i], ocr_text);
     CHECK(strcmp(gt_text_[i].c_str(), ocr_text.c_str()) == 0) << "Failed with lang = " << langs_[i];
   }
 }

unittest/include_gunit.h

@@ -20,6 +20,19 @@
 
 const char *FLAGS_test_tmpdir = "./tmp";
 
+namespace tesseract {
+
+void trim(std::string &s) {
+  s.erase(s.begin(), std::find_if(s.begin(), s.end(), [](unsigned char ch) {
+    return !std::isspace(ch);
+  }));
+  s.erase(std::find_if(s.rbegin(), s.rend(), [](unsigned char ch) {
+    return !std::isspace(ch);
+  }).base(), s.end());
+}
+
+} // namespace tesseract
+
 class file : public tesseract::File {
 public:
   static void MakeTmpdir() {