replace boost string functions

split/join
trim
starts_with/ends_with

src/rime/algo/calculus.cc

@@ -4,9 +4,9 @@
 //
 // 2012-01-17 GONG Chen <[email protected]>
 //
-#include <boost/algorithm/string.hpp>
 #include <utf8.h>
 #include <rime/algo/calculus.h>
+#include <rime/algo/strings.h>
 #include <rime/common.h>
 
 namespace rime {
@@ -31,9 +31,8 @@ Calculation* Calculus::Parse(const string& definition) {
   size_t sep = definition.find_first_not_of("zyxwvutsrqponmlkjihgfedcba");
   if (sep == string::npos)
     return NULL;
-  vector<string> args;
-  boost::split(args, definition,
-               boost::is_from_range(definition[sep], definition[sep]));
+  const string& delim = definition.substr(sep, 1);
+  vector<string> args = strings::split(definition, delim);
   if (args.empty())
     return NULL;
   auto it = factories_.find(args[0]);

src/rime/algo/strings.h

@@ -35,6 +35,32 @@ inline string join(std::initializer_list<C>&& container, T&& delim) {
   return join(std::begin(container), std::end(container), delim);
 }
 
+inline void trim_left(string& str) {
+  str.erase(str.begin(),
+            std::find_if(str.begin(), str.end(),
+                         [](unsigned char ch) { return !std::isspace(ch); }));
+}
+
+inline void trim_right(string& str) {
+  str.erase(std::find_if(str.rbegin(), str.rend(),
+                         [](unsigned char ch) { return !std::isspace(ch); })
+                .base(),
+            str.end());
+}
+
+inline void trim(string& str) {
+  trim_right(str);
+  trim_left(str);
+}
+
+inline bool starts_with(const string& str, const string& prefix) {
+  return str.rfind(prefix, 0) == 0;
+}
+
+inline bool ends_with(const string& str, const string& suffix) {
+  return str.find(suffix, str.length() - suffix.length()) != std::string::npos;
+}
+
 }  // namespace strings
 }  // namespace rime
 

src/rime/config/config_data.cc

@@ -6,9 +6,9 @@
 #include <cstdlib>
 #include <fstream>
 #include <sstream>
-#include <boost/algorithm/string.hpp>
 #include <filesystem>
 #include <yaml-cpp/yaml.h>
+#include <rime/algo/strings.h>
 #include <rime/config/config_compiler.h>
 #include <rime/config/config_cow_ref.h>
 #include <rime/config/config_data.h>
@@ -207,15 +207,13 @@ bool ConfigData::TraverseWrite(const string& path, an<ConfigItem> item) {
 }
 
 vector<string> ConfigData::SplitPath(const string& path) {
-  vector<string> keys;
-  auto is_separator = boost::is_any_of("/");
-  auto trimmed_path = boost::trim_left_copy_if(path, is_separator);
-  boost::split(keys, trimmed_path, is_separator);
+  auto trimmed_path = path.substr(strings::starts_with(path, "/"));
+  vector<string> keys = strings::split(trimmed_path, "/");
   return keys;
 }
 
 string ConfigData::JoinPath(const vector<string>& keys) {
-  return boost::join(keys, "/");
+  return strings::join(keys, "/");
 }
 
 an<ConfigItem> ConfigData::Traverse(const string& path) {

src/rime/dict/reverse_lookup_dictionary.cc

@@ -8,12 +8,12 @@
 #include <cfloat>
 #include <cstdlib>
 #include <sstream>
-#include <boost/algorithm/string.hpp>
 #include <filesystem>
 #include <rime/resource.h>
 #include <rime/schema.h>
 #include <rime/service.h>
 #include <rime/ticket.h>
+#include <rime/algo/strings.h>
 #include <rime/dict/dict_settings.h>
 #include <rime/dict/reverse_lookup_dictionary.h>
 
@@ -107,15 +107,15 @@ bool ReverseDb::Build(DictSettings* settings,
   // save reverse lookup entries
   for (const auto& v : rev_table) {
     const string& key(v.first);
-    string value(boost::algorithm::join(v.second, " "));
+    string value(strings::join(v.second, " "));
     key_trie_builder.Add(key, 0.0, &key_ids[i]);
     value_trie_builder.Add(value, 0.0, &value_ids[i]);
     ++i;
   }
   // save stems
   for (const auto& v : stems) {
     string key(v.first + kStemKeySuffix);
-    string value(boost::algorithm::join(v.second, " "));
+    string value(strings::join(v.second, " "));
     key_trie_builder.Add(key, 0.0, &key_ids[i]);
     value_trie_builder.Add(value, 0.0, &value_ids[i]);
     ++i;

src/rime/dict/table_db.cc

@@ -4,7 +4,7 @@
 //
 // 2013-04-18 GONG Chen <[email protected]>
 //
-#include <boost/algorithm/string.hpp>
+#include <rime/algo/strings.h>
 #include <rime/dict/table_db.h>
 #include <rime/dict/user_db.h>
 
@@ -22,7 +22,7 @@ static bool rime_table_entry_parser(const Tsv& row,
     return false;
   }
   string code(row[1]);
-  boost::algorithm::trim(code);
+  strings::trim(code);
   *key = code + " \t" + row[0];
   UserDbValue v;
   if (row.size() >= 3 && !row[2].empty()) {
@@ -42,13 +42,13 @@ static bool rime_table_entry_formatter(const string& key,
                                        Tsv* tsv) {
   Tsv& row(*tsv);
   // key ::= code <space> <Tab> phrase
-  boost::algorithm::split(row, key, boost::algorithm::is_any_of("\t"));
+  row = strings::split(key, "\t");
   if (row.size() != 2 || row[0].empty() || row[1].empty())
     return false;
   UserDbValue v(value);
   if (v.commits < 0)  // deleted entry
     return false;
-  boost::algorithm::trim(row[0]);  // remove trailing space
+  strings::trim(row[0]);  // remove trailing space
   row[0].swap(row[1]);
   row.push_back(std::to_string(v.commits));
   return true;

src/rime/dict/tsv.cc

@@ -5,7 +5,7 @@
 // 2013-04-14 GONG Chen <[email protected]>
 //
 #include <fstream>
-#include <boost/algorithm/string.hpp>
+#include <rime/algo/strings.h>
 #include <rime/common.h>
 #include <rime/dict/db_utils.h>
 #include <rime/dict/tsv.h>
@@ -24,15 +24,15 @@ int TsvReader::operator()(Sink* sink) {
   bool enable_comment = true;
   while (getline(fin, line)) {
     ++line_no;
-    boost::algorithm::trim_right(line);
+    strings::trim_right(line);
     // skip empty lines and comments
     if (line.empty())
       continue;
     if (enable_comment && line[0] == '#') {
-      if (boost::starts_with(line, "#@")) {
+      if (strings::starts_with(line, "#@")) {
         // metadata
         line.erase(0, 2);
-        boost::algorithm::split(row, line, boost::algorithm::is_any_of("\t"));
+        row = strings::split(line, "\t");
         if (row.size() != 2 || !sink->MetaPut(row[0], row[1])) {
           LOG(WARNING) << "invalid metadata at line " << line_no << ".";
         }
@@ -43,7 +43,7 @@ int TsvReader::operator()(Sink* sink) {
       continue;
     }
     // read a tsv entry
-    boost::algorithm::split(row, line, boost::algorithm::is_any_of("\t"));
+    row = strings::split(line, "\t");
     if (!parser_(row, &key, &value) || !sink->Put(key, value)) {
       LOG(WARNING) << "invalid entry at line " << line_no << ".";
       continue;

src/rime/dict/user_db.cc

@@ -6,7 +6,7 @@
 //
 #include <cstdlib>
 #include <sstream>
-#include <boost/algorithm/string.hpp>
+#include <rime/algo/strings.h>
 #include <rime/service.h>
 #include <rime/algo/dynamics.h>
 #include <rime/dict/text_db.h>
@@ -25,8 +25,7 @@ string UserDbValue::Pack() const {
 }
 
 bool UserDbValue::Unpack(const string& value) {
-  vector<string> kv;
-  boost::split(kv, value, boost::is_any_of(" "));
+  vector<string> kv = strings::split(value, " ");
   for (const string& k_eq_v : kv) {
     size_t eq = k_eq_v.find('=');
     if (eq == string::npos)
@@ -83,7 +82,7 @@ static bool userdb_entry_formatter(const string& key,
                                    const string& value,
                                    Tsv* tsv) {
   Tsv& row(*tsv);
-  boost::algorithm::split(row, key, boost::algorithm::is_any_of("\t"));
+  row = strings::split(key, "\t");
   if (row.size() != 2 || row[0].empty() || row[1].empty())
     return false;
   row.push_back(value);
@@ -107,7 +106,7 @@ bool UserDbHelper::UpdateUserInfo() {
 }
 
 bool UserDbHelper::IsUniformFormat(const string& file_name) {
-  return boost::ends_with(file_name, plain_userdb_extension);
+  return strings::ends_with(file_name, plain_userdb_extension);
 }
 
 bool UserDbHelper::UniformBackup(const string& snapshot_file) {
@@ -147,10 +146,10 @@ string UserDbHelper::GetDbName() {
   string name;
   if (!db_->MetaFetch("/db_name", &name))
     return name;
-  auto ext = boost::find_last(name, ".userdb");
-  if (!ext.empty()) {
+  auto ext = name.rfind(".userdb");
+  if (ext != std::string::npos) {
     // remove ".userdb.*"
-    name.erase(ext.begin(), name.end());
+    name.erase(ext, name.length() - ext);
   }
   return name;
 }

src/rime/gear/script_translator.cc

@@ -9,7 +9,6 @@
 #include <algorithm>
 #include <stack>
 #include <cmath>
-#include <boost/algorithm/string/join.hpp>
 #include <boost/range/adaptor/reversed.hpp>
 #include <rime/composition.h>
 #include <rime/candidate.h>
@@ -18,6 +17,7 @@
 #include <rime/engine.h>
 #include <rime/schema.h>
 #include <rime/translation.h>
+#include <rime/algo/strings.h>
 #include <rime/algo/syllabifier.h>
 #include <rime/dict/corrector.h>
 #include <rime/dict/dictionary.h>
@@ -214,7 +214,7 @@ string ScriptTranslator::Spell(const Code& code) {
   vector<string> syllables;
   if (!dict_ || !dict_->Decode(code, &syllables) || syllables.empty())
     return result;
-  result = boost::algorithm::join(syllables, string(1, delimiters_.at(0)));
+  result = strings::join(syllables, string(1, delimiters_.at(0)));
   comment_formatter_.Apply(&result);
   return result;
 }

src/rime/gear/switch_translator.cc

@@ -4,7 +4,7 @@
 //
 // 2013-05-26 GONG Chen <[email protected]>
 //
-#include <boost/algorithm/string.hpp>
+#include <rime/algo/strings.h>
 #include <rime/candidate.h>
 #include <rime/common.h>
 #include <rime/config.h>
@@ -184,7 +184,7 @@ void FoldedOptions::Append(const SwitchOption& option, size_t state_index) {
 }
 
 void FoldedOptions::Finish() {
-  text_ = prefix_ + boost::algorithm::join(labels_, separator_) + suffix_;
+  text_ = prefix_ + strings::join(labels_, separator_) + suffix_;
 }
 
 class SwitchTranslation : public FifoTranslation {

src/rime/gear/unity_table_encoder.cc

@@ -4,7 +4,7 @@
 //
 // 2013-08-31 GONG Chen <[email protected]>
 //
-#include <boost/algorithm/string.hpp>
+#include <rime/algo/strings.h>
 #include <rime/dict/dict_settings.h>
 #include <rime/dict/user_dictionary.h>
 #include <rime/dict/reverse_lookup_dictionary.h>
@@ -62,7 +62,7 @@ bool UnityTableEncoder::TranslateWord(const string& word,
   string str_list;
   if (rev_dict_->LookupStems(word, &str_list) ||
       rev_dict_->ReverseLookup(word, &str_list)) {
-    boost::split(*code, str_list, boost::is_any_of(" "));
+    *code = strings::split(str_list, " ");
     return !code->empty();
   }
   return false;
@@ -80,7 +80,7 @@ size_t UnityTableEncoder::LookupPhrases(UserDictEntryIterator* result,
 }
 
 bool UnityTableEncoder::HasPrefix(const string& key) {
-  return boost::starts_with(key, kEncodedPrefix);
+  return strings::starts_with(key, kEncodedPrefix);
 }
 
 bool UnityTableEncoder::AddPrefix(string* key) {