Refresh README with product-focused overview

Author: nufegia

README.md

@@ -29,7 +29,7 @@ keywords = [
   "sprite",
 ]
 classifiers = [
-  "Development Status :: 4 - Beta",
+  "Development Status :: 5 - Production/Stable",
   "Environment :: Console",
   "Intended Audience :: Science/Research",
   "Programming Language :: Python :: 3",

pyproject.toml

@@ -61,9 +61,6 @@ class OrdinalConfig:
         "stage",
         "likert",
         "rating",
-        "评分",
-        "等级",
-        "分级",
     )
 
 

src/pcr_audit/config.py

@@ -23,18 +23,18 @@
 # ---------------------------------------------------------------------------
 
 COLUMN_PATTERNS: dict[str, list[str]] = {
-    "N":        [r"^n$", r"^sample[_ ]?size$", r"^cases?$", r"^number$", r"samplesize", r"样本量", r"例数", r"人数"],
-    "Mean":     [r"^mean$", r"^means$", r"^average$", r"均值", r"均数", r"平均值", r"平均数"],
-    "SD":       [r"^sd$", r"^std$", r"^stdev$", r"标准差", r"stdev"],
-    "SE":       [r"^se$", r"^sem$", r"^standard[_ ]?error$", r"标准误", r"sterr"],
-    "CI_low":   [r"cilow", r"cilower", r"ci[_\- ]?l", r"lcl", r"^lower$", r"置信区间下限", r"下限", r"^low$"],
-    "CI_high":  [r"cihigh", r"ciupper", r"ci[_\- ]?u", r"ucl", r"^upper$", r"置信区间上限", r"上限", r"^high$"],
-    "count":    [r"^count$", r"n[_\- ]?pos", r"^freq$", r"频数", r"计数"],
-    "percent":  [r"percent", r"percentage", r"^prop$", r"^rate$", r"百分比", r"比例", r"率"],
-    "effect":   [r"^or$", r"oddsratio", r"^rr$", r"riskratio", r"^hr$", r"hazardratio", r"效应量", r"比值比", r"风险比"],
-    "t":        [r"^t$", r"t[_\- ]?value", r"t[_\- ]?stat", r"t统计"],
-    "df":       [r"^df$", r"^dof$", r"自由度", r"degreeoffreedom"],
-    "p":        [r"^p$", r"p[_\- ]?value", r"^pval$", r"p值"],
+    "N":        [r"^n$", r"^sample[_ ]?size$", r"^cases?$", r"^number$", r"samplesize"],
+    "Mean":     [r"^mean$", r"^means$", r"^average$"],
+    "SD":       [r"^sd$", r"^std$", r"^stdev$"],
+    "SE":       [r"^se$", r"^sem$", r"^standard[_ ]?error$", r"sterr"],
+    "CI_low":   [r"cilow", r"cilower", r"ci[_\- ]?l", r"lcl", r"^lower$", r"^low$"],
+    "CI_high":  [r"cihigh", r"ciupper", r"ci[_\- ]?u", r"ucl", r"^upper$", r"^high$"],
+    "count":    [r"^count$", r"n[_\- ]?pos", r"^freq$"],
+    "percent":  [r"percent", r"percentage", r"^prop$", r"^rate$"],
+    "effect":   [r"^or$", r"oddsratio", r"^rr$", r"riskratio", r"^hr$", r"hazardratio"],
+    "t":        [r"^t$", r"t[_\- ]?value", r"t[_\- ]?stat"],
+    "df":       [r"^df$", r"^dof$", r"degreeoffreedom"],
+    "p":        [r"^p$", r"p[_\- ]?value", r"^pval$"],
 }
 
 

src/pcr_audit/crosscheck.py

@@ -18,12 +18,12 @@
 
 
 STAT_NAMES = {
-    "n": ("n", "sample_size", "cases", "samplesize", "number", "样本量", "例数", "人数"),
-    "mean": ("mean", "means", "average", "均值", "均数", "平均值", "平均数"),
-    "sd": ("sd", "std", "stdev", "标准差"),
-    "se": ("se", "sem", "standard_error", "标准误", "sterr"),
-    "count": ("count", "freq", "frequency", "n_pos", "频数", "计数"),
-    "percent": ("percent", "percentage", "prop", "rate", "百分比", "比例", "率"),
+    "n": ("n", "sample_size", "cases", "samplesize", "number"),
+    "mean": ("mean", "means", "average"),
+    "sd": ("sd", "std", "stdev"),
+    "se": ("se", "sem", "standard_error", "sterr"),
+    "count": ("count", "freq", "frequency", "n_pos"),
+    "percent": ("percent", "percentage", "prop", "rate"),
 }
 
 

src/pcr_audit/data_trace.py

@@ -40,7 +40,7 @@ def _p_columns(df: pd.DataFrame) -> list[str]:
     for col in df.columns:
         normalized = _normalize_name(col)
         raw = str(col).lower()
-        if re.search(r"^p$", normalized) or re.search(r"pvalue", normalized) or re.search(r"p[_ ]?value", raw) or re.search(r"^pval", normalized) or re.search(r"p值", raw):
+        if re.search(r"^p$", normalized) or re.search(r"pvalue", normalized) or re.search(r"p[_ ]?value", raw) or re.search(r"^pval", normalized):
             columns.append(str(col))
     return columns
 

src/pcr_audit/detectors/p_values.py

@@ -98,7 +98,7 @@ def enrich_finding_explanation(finding: Finding) -> None:
     if not finding.location:
         finding.location = finding.table
     if not finding.review_actions:
-        finding.review_actions = finding.review_steps
+        finding.review_actions = finding.review_steps or finding.suggestion
     if not finding.confidence_basis:
         finding.confidence_basis = (
             "Generated from deterministic rules or reproducible formulas; still requires human judgment considering study design, original records, and material extraction quality."

src/pcr_audit/detectors/r/adapters.py

@@ -38,10 +38,10 @@
     "man-made brainpower": "artificial intelligence",
 }
 
-TEXT_TOKEN_RE = re.compile(r"[A-Za-z0-9\u4e00-\u9fff]{2,}")
+TEXT_TOKEN_RE = re.compile(r"[A-Za-z0-9]{2,}")
 EMAIL_RE = re.compile(r"\b[A-Z0-9._%+-]+@([A-Z0-9.-]+\.[A-Z]{2,})\b", re.I)
-AUTHOR_LINE_RE = re.compile(r"^\s*(?:authors?|作者)\s*[:：]\s*(?P<value>.+)$", re.I | re.M)
-INSTITUTION_LINE_RE = re.compile(r"^\s*(?:affiliations?|institutions?|单位|机构)\s*[:：]\s*(?P<value>.+)$", re.I | re.M)
+AUTHOR_LINE_RE = re.compile(r"^\s*(?:authors?)\s*[:：]\s*(?P<value>.+)$", re.I | re.M)
+INSTITUTION_LINE_RE = re.compile(r"^\s*(?:affiliations?|institutions?)\s*[:：]\s*(?P<value>.+)$", re.I | re.M)
 
 
 @dataclass