1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/browser/components/translation/cld2/internal/compact_lang_det_hint_code.h Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,95 @@
1.4 +// Copyright 2013 Google Inc. All Rights Reserved.
1.5 +//
1.6 +// Licensed under the Apache License, Version 2.0 (the "License");
1.7 +// you may not use this file except in compliance with the License.
1.8 +// You may obtain a copy of the License at
1.9 +//
1.10 +// http://www.apache.org/licenses/LICENSE-2.0
1.11 +//
1.12 +// Unless required by applicable law or agreed to in writing, software
1.13 +// distributed under the License is distributed on an "AS IS" BASIS,
1.14 +// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
1.15 +// See the License for the specific language governing permissions and
1.16 +// limitations under the License.
1.17 +
1.18 +//
1.19 +// Author: dsites@google.com (Dick Sites)
1.20 +//
1.21 +
1.22 +#ifndef I18N_ENCODINGS_COMPACT_LANG_DET_COMPACT_LANG_DET_HINT_CODE_H__
1.23 +#define I18N_ENCODINGS_COMPACT_LANG_DET_COMPACT_LANG_DET_HINT_CODE_H__
1.24 +
1.25 +
1.26 +#include <string>
1.27 +#include "integral_types.h"
1.28 +#include "lang_script.h"
1.29 +#include "../public/encodings.h"
1.30 +
1.31 +namespace CLD2 {
1.32 +
1.33 +// Packed <Language, weight>, weight in [-32..31] (powers of 2**1.6 ~=3.03)
1.34 +// Full language in bottom 10 bits, weight in top 6 bits
1.35 +typedef int16 OneCLDLangPrior;
1.36 +
1.37 +const int kMaxOneCLDLangPrior = 14;
1.38 +typedef struct {
1.39 + int32 n;
1.40 + OneCLDLangPrior prior[kMaxOneCLDLangPrior];
1.41 +} CLDLangPriors;
1.42 +
1.43 +// Reading exposed here; setting hidden in .cc
1.44 +inline int GetCLDPriorWeight(OneCLDLangPrior olp) {
1.45 + return olp >> 10;
1.46 +}
1.47 +inline Language GetCLDPriorLang(OneCLDLangPrior olp) {
1.48 + return static_cast<Language>(olp & 0x3ff);
1.49 +}
1.50 +
1.51 +inline int32 GetCLDLangPriorCount(CLDLangPriors* lps) {
1.52 + return lps->n;
1.53 +}
1.54 +
1.55 +inline void InitCLDLangPriors(CLDLangPriors* lps) {
1.56 + lps->n = 0;
1.57 +}
1.58 +
1.59 +// Trim language priors to no more than max_entries, keeping largest abs weights
1.60 +void TrimCLDLangPriors(int max_entries, CLDLangPriors* lps);
1.61 +
1.62 +// Trim language tag string to canonical form for each language
1.63 +// Input is from GetLangTagsFromHtml(), already lowercased
1.64 +std::string TrimCLDLangTagsHint(const std::string& langtags);
1.65 +
1.66 +// Add hints to vector of langpriors
1.67 +// Input is from GetLangTagsFromHtml(), already lowercased
1.68 +void SetCLDLangTagsHint(const std::string& langtags, CLDLangPriors* langpriors);
1.69 +
1.70 +// Add hints to vector of langpriors
1.71 +// Input is from HTTP content-language
1.72 +void SetCLDContentLangHint(const char* contentlang, CLDLangPriors* langpriors);
1.73 +
1.74 +// Add hints to vector of langpriors
1.75 +// Input is from GetTLD(), already lowercased
1.76 +void SetCLDTLDHint(const char* tld, CLDLangPriors* langpriors);
1.77 +
1.78 +// Add hints to vector of langpriors
1.79 +// Input is from DetectEncoding()
1.80 +void SetCLDEncodingHint(Encoding enc, CLDLangPriors* langpriors);
1.81 +
1.82 +// Add hints to vector of langpriors
1.83 +// Input is from random source
1.84 +void SetCLDLanguageHint(Language lang, CLDLangPriors* langpriors);
1.85 +
1.86 +// Make printable string of priors
1.87 +std::string DumpCLDLangPriors(const CLDLangPriors* langpriors);
1.88 +
1.89 +
1.90 +// Get language tag hints from HTML body
1.91 +// Normalize: remove spaces and make lowercase comma list
1.92 +std::string GetLangTagsFromHtml(const char* utf8_body, int32 utf8_body_len,
1.93 + int32 max_scan_bytes);
1.94 +
1.95 +} // End namespace CLD2
1.96 +
1.97 +#endif // I18N_ENCODINGS_COMPACT_LANG_DET_COMPACT_LANG_DET_HINT_CODE_H__
1.98 +
