Files
Jia Suiandfanlusky 60c2b26698 fix(engine,server): 全拼纠错收窄「多字」纠错并让开关真正关闭纠错 (#889)
* fix(engine,server): 全拼多字纠错收窄为仅 QWERTY 邻键插入

多字(插入)纠错只保留「误插 QWERTY 邻键」:生成器 insertion_variants 不再生成重复相邻键(双击),生成式空间不再生成任意字母插入(远键插入)。autocorrect_table.h 随生成器重新生成。

* fix(engine): 全拼纠错开关关闭时真正不再改写切分

纠错开关关掉后真正不纠错:把手误拼写别名层从常开改为跟掩码走(ü 归一 jv->ju / nue->nve 仍常开);两个早于掩码可知的切分入口 QuanpinScheme::build_request 与 QuanpinDictionary::get_or_compute_segments 显式带上掩码,切分缓存键随之隔离。

* fix(engine,server): 纠错开关关闭时辅助码、写回与排序上下文也按字面切分

关掉全拼纠错后,只有方案的 build_request 和查询自身切分不再跑手误别名层,
其余几条路径仍用默认(带别名)的切分,和查询落在不同的键上:

- 辅助码路径自己切底串作为显式切分交给词典,haoyonga 仍被切成 hao'yong'ga;
- 调频、造词、删词、建 SQL、校验去掉分隔符后重切,开关关掉时选中 hao'yong'a
  的词会去更新 hao'yong'ga,调频静默失败;
- 排序 / 固定位置的上下文 key(引擎与 server 两处)仍按别名切,laing 与 liang
  的固定位置互相串。

改动:
- cut_pinyin_by_mode 的 include_correction_aliases 改为必填,每个调用点显式决定;
  双拼词典、设置页词库校验和离线工具传 true,保持原行为。
- 新增 quanpin::correction_aliases_enabled,方案、词典、辅助码与排序 key 共用
  (只看换位、邻键两位),修正只带删除 / 插入位的掩码在方案与词典两层判断不一致。
- 词典按最近一次 query 的掩码记录别名开关,写回路径统一走 cut_for_write。
- 会话改为在换方案、改开关时 bind_quanpin_scheme 一次,不再每次 refresh 下发。
- 方案只存 bool;jv -> ju 放回纠错表,去掉合并表。
- 在 CI 实际执行的 test_input_session 中补回归用例(写回、辅助码、固定位置、
  方案掩码),每条都确认回退修复后会失败。

---------

Co-authored-by: fanlusky <1730976608@qq.com>
2026-10-12 20:41:23 +08:00

162 lines
4.3 KiB
C++

#include "quanpin_scheme.h"
#include "../common/helpcode_utils.h"
#include "../quanpin/quanpin_query.h"
#include "../quanpin/quanpin_utils.h"
#include "../shuangpin/shuangpin_query.h"
#include <cctype>
namespace
{
bool is_alpha_vk(ImeKeyCode vk)
{
return vk >= 'A' && vk <= 'Z';
}
} // namespace
void QuanpinScheme::reset()
{
raw_input_.clear();
key_strokes_.clear();
}
void QuanpinScheme::set_raw_input(const std::string &raw_input, const std::string &raw_input_with_cases)
{
raw_input_ = raw_input_with_cases.empty() ? raw_input : raw_input_with_cases;
key_strokes_.clear();
}
void QuanpinScheme::set_autocorrect_types(unsigned autocorrect_types)
{
include_correction_aliases_ = quanpin::correction_aliases_enabled(autocorrect_types);
}
void QuanpinScheme::handle_key(ImeKeyCode vk, ImeModifierMask modifiers_down, ImeCharacter wch)
{
if (vk == ImeKey::Backspace)
{
if (!raw_input_.empty())
{
raw_input_.pop_back();
}
if (!key_strokes_.empty())
{
key_strokes_.pop_back();
}
return;
}
if (vk == ImeKey::Escape || vk == ImeKey::Return)
{
reset();
return;
}
if (vk == ImeKey::Apostrophe)
{
if (raw_input_.empty() || raw_input_.back() != '\'')
{
key_strokes_.push_back(KeyStroke{vk, modifiers_down, wch});
raw_input_.push_back('\'');
}
return;
}
if (!is_alpha_vk(vk))
{
return;
}
key_strokes_.push_back(KeyStroke{vk, modifiers_down, wch});
if (wch >= L'A' && wch <= L'Z')
{
raw_input_.push_back(static_cast<char>(wch));
}
else if (wch >= L'a' && wch <= L'z')
{
raw_input_.push_back(static_cast<char>(wch));
}
else
{
raw_input_.push_back(static_cast<char>(vk + ('a' - 'A')));
}
}
QueryRequest QuanpinScheme::build_request() const
{
QueryRequest request;
request.scheme = type();
request.raw_input_with_cases = raw_input_;
request.raw_input.reserve(raw_input_.size());
for (const char ch : raw_input_)
{
request.raw_input.push_back(ch == '\'' ? ch : static_cast<char>(std::tolower(static_cast<unsigned char>(ch))));
}
request.key_strokes = key_strokes_;
const size_t helpcode_length =
quanpin::detect_active_helpcode_length(request.raw_input, request.raw_input_with_cases);
std::string normalized_source = quanpin::strip_active_helpcodes(request.raw_input, request.raw_input_with_cases);
request.normalized_input.reserve(normalized_source.size());
for (const char ch : normalized_source)
{
if (ch != '\'')
{
request.normalized_input.push_back(ch);
}
}
if (!request.normalized_input.empty())
{
try
{
const auto segments =
quanpin::cut_pinyin_by_mode(normalized_source, "correction", include_correction_aliases_);
if (!segments.empty())
{
request.normalized_segmentation = quanpin::join_segments(segments.front());
}
}
catch (...)
{
}
}
if (request.normalized_segmentation.empty())
{
request.normalized_segmentation = normalized_source;
}
std::string cased_source = request.raw_input_with_cases;
if (helpcode_length > 0 && cased_source.size() >= helpcode_length)
{
cased_source = cased_source.substr(0, cased_source.size() - helpcode_length);
}
request.raw_segmentation = shuangpin::apply_segmentation_cases(request.normalized_segmentation, cased_source);
if (!cased_source.empty() && cased_source.back() == '\'' &&
(request.raw_segmentation.empty() || request.raw_segmentation.back() != '\''))
{
request.raw_segmentation.push_back('\'');
}
if (helpcode_length > 0)
{
request.raw_segmentation += "'";
request.raw_segmentation +=
request.raw_input_with_cases.substr(request.raw_input_with_cases.size() - helpcode_length, helpcode_length);
}
request.segmentation = request.normalized_segmentation;
request.valid = !request.normalized_input.empty();
return request;
}
std::string QuanpinScheme::get_preedit() const
{
return raw_input_;
}
SchemeType QuanpinScheme::type() const
{
return SchemeType::Quanpin;
}