execsql2 2.22.4__py3-none-any.whl → 2.22.5__py3-none-any.whl

execsql/format.py

@@ -14,6 +14,7 @@ from __future__ import annotations
 import contextlib
 import io
 import re
+from collections.abc import Iterator
 from pathlib import Path
 
 __all__ = ["collect_paths", "format_file", "main", "parse_keyword"]
@@ -169,6 +170,170 @@ def parse_keyword(payload: str) -> str:
 _EXECSQL_VAR_RE = re.compile(r"""!(['"]?)!([^!\s][^!]*)!\1!|!\{[^}]+\}!""")
 
 
+# Characters that may appear in an identifier, used to tell an ``E'...'``
+# escape-string prefix from a trailing ``e`` on an identifier (``abce'x'`` is
+# the identifier ``abce`` followed by a separate literal).
+_IDENT_CHAR_RE = re.compile(r"[A-Za-z0-9_$]")
+
+
+def _iter_sql_literals(sql: str) -> Iterator[tuple[int, int, str]]:
+    """Yield ``(start, end, kind)`` spans for every string literal in *sql*.
+
+    This is a lexical walk rather than a regex sweep, because an apostrophe
+    only starts a literal in some contexts.  Line comments, block comments
+    (which nest in PostgreSQL), and quoted identifiers are skipped entirely,
+    so quotes inside them never open a literal.
+
+    *kind* is one of:
+
+    ``"estring"``
+        A PostgreSQL escape string, ``E'...'``.  Backslash escapes are
+        significant inside it, which is what makes it unsafe to hand to
+        sqlglot — see :func:`_protect_estrings`.
+    ``"plain"``
+        An ordinary ``'...'`` literal, where only a doubled ``''`` escapes.
+    ``"dollar"``
+        A dollar-quoted body, ``$$...$$`` or ``$tag$...$tag$``, in which no
+        character is special.
+    """
+    i, n = 0, len(sql)
+    while i < n:
+        c = sql[i]
+
+        if c == "-" and sql.startswith("--", i):
+            nl = sql.find("\n", i)
+            i = n if nl < 0 else nl + 1
+            continue
+
+        if c == "/" and sql.startswith("/*", i):
+            depth, i = 1, i + 2
+            while i < n and depth:
+                if sql.startswith("/*", i):
+                    depth, i = depth + 1, i + 2
+                elif sql.startswith("*/", i):
+                    depth, i = depth - 1, i + 2
+                else:
+                    i += 1
+            continue
+
+        if c == '"':
+            i += 1
+            while i < n:
+                if sql[i] == '"':
+                    if sql.startswith('""', i):
+                        i += 2
+                        continue
+                    i += 1
+                    break
+                i += 1
+            continue
+
+        if c == "$":
+            m = _DOLLAR_QUOTE_RE.match(sql, i)
+            if m:
+                tag = m.group(0)
+                close = sql.find(tag, m.end())
+                stop = n if close < 0 else close + len(tag)
+                yield i, stop, "dollar"
+                i = stop
+                continue
+
+        # E'...' — only when the `e` is not the tail of an identifier.
+        if c in "eE" and sql.startswith("'", i + 1) and not (i and _IDENT_CHAR_RE.match(sql[i - 1])):
+            j = i + 2
+            while j < n:
+                if sql[j] == "\\" and j + 1 < n:
+                    j += 2
+                    continue
+                if sql[j] == "'":
+                    if sql.startswith("''", j):
+                        j += 2
+                        continue
+                    j += 1
+                    break
+                j += 1
+            yield i, j, "estring"
+            i = j
+            continue
+
+        if c == "'":
+            j = i + 1
+            while j < n:
+                if sql[j] == "'":
+                    if sql.startswith("''", j):
+                        j += 2
+                        continue
+                    j += 1
+                    break
+                j += 1
+            yield i, j, "plain"
+            i = j
+            continue
+
+        i += 1
+
+
+def _sql_literal_texts(sql: str) -> list[str]:
+    """Return the source text of every string literal in *sql*, in order."""
+    return [sql[start:end] for start, end, _ in _iter_sql_literals(sql)]
+
+
+# Placeholder standing in for a masked E'...' literal.  It is a plain literal
+# so that sqlglot sees a string where a string was, and lays the statement out
+# the same way it would have for the real one.
+_ESTR_PLACEHOLDER = "execsqlestr"
+_ESTR_RESTORE_RE = re.compile(rf"'{_ESTR_PLACEHOLDER}(\d+)'")
+
+
+def _protect_estrings(sql: str) -> tuple[str, list[str]]:
+    """Replace ``E'...'`` literals with placeholders, return (protected, originals).
+
+    sqlglot's Postgres tokenizer consumes backslash escapes inside an escape
+    string but its generator never re-emits them, so ``E'\\\\s+'`` round-trips
+    to ``e'\\s+'`` — a different string (tobymao/sqlglot#8191).  Masking these
+    literals keeps them out of the round trip entirely and restores them
+    byte-for-byte, independently of the installed sqlglot version.
+    """
+    originals: list[str] = []
+    out: list[str] = []
+    last = 0
+    for start, end, kind in _iter_sql_literals(sql):
+        if kind != "estring":
+            continue
+        out.append(sql[last:start])
+        out.append(f"'{_ESTR_PLACEHOLDER}{len(originals)}'")
+        originals.append(sql[start:end])
+        last = end
+    if not originals:
+        return sql, []
+    out.append(sql[last:])
+    return "".join(out), originals
+
+
+def _restore_estrings(sql: str, originals: list[str]) -> str | None:
+    """Put masked ``E'...'`` literals back; return None if any went missing.
+
+    A missing placeholder means sqlglot dropped or mangled the literal, so the
+    caller must fall back to the unformatted text rather than emit SQL with a
+    placeholder left in it.
+    """
+    if not originals:
+        return sql
+    seen: set[int] = set()
+
+    def replace(m: re.Match[str]) -> str:
+        idx = int(m.group(1))
+        if idx >= len(originals):
+            return m.group(0)
+        seen.add(idx)
+        return originals[idx]
+
+    restored = _ESTR_RESTORE_RE.sub(replace, sql)
+    if len(seen) != len(originals) or _ESTR_PLACEHOLDER in restored:
+        return None
+    return restored
+
+
 def _protect_variables(sql: str) -> tuple[str, list[tuple[str, str]]]:
     """Replace execsql substitutions with valid SQL identifiers, return (protected, replacements)."""
     replacements: list[tuple[str, str]] = []
@@ -224,6 +389,9 @@ def _sqlglot_format(
 
     text = "\n".join(sql_lines)
     protected, replacements = _protect_variables(text)
+    # Keep E'...' literals out of the round trip; sqlglot loses their
+    # backslash escapes (see _protect_estrings).
+    protected, estrings = _protect_estrings(protected)
 
     # Count semicolons in input as a rough statement count.
     input_semis = protected.count(";")
@@ -275,8 +443,29 @@ def _sqlglot_format(
         output_alnum_len = len(_alnum.sub("", joined))
         if input_alnum_len and output_alnum_len < input_alnum_len * 0.7:
             return sql_lines
+
+        # Literal round-trip check.  The alphanumeric check above is a
+        # gross-token-loss detector: it deletes every non-alphanumeric
+        # character before counting, so a transformation that changes only
+        # punctuation — a dropped backslash, a swapped quote — is invisible to
+        # it at any threshold.  Requiring every literal to survive verbatim
+        # closes that blind spot for the whole class, not just for the E'...'
+        # case masked above.
+        #
+        # Containment, not equality: sqlglot rewrites dialect-specific
+        # constructs into Postgres equivalents that can legitimately repeat a
+        # literal (``SUBSTRING_INDEX(d, ',', 1)`` expands into a form using
+        # ``','`` more than once), so the output may hold more literals than
+        # the input.  What must never happen is an input literal coming back
+        # altered or not at all — that is content loss, and it falls back.
+        if set(_sql_literal_texts(protected)) - set(_sql_literal_texts(joined)):
+            return sql_lines
+
         joined = re.sub(r"\bINTO TEMPORARY\b(?!\s+TABLE)", "INTO TEMPORARY TABLE", joined)
-        return _restore_variables(joined, replacements).split("\n")
+        restored = _restore_estrings(joined, estrings)
+        if restored is None:
+            return sql_lines
+        return _restore_variables(restored, replacements).split("\n")
     except Exception:
         return sql_lines
 

{execsql2-2.22.4.dist-info → execsql2-2.22.5.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: execsql2
-Version: 2.22.4
+Version: 2.22.5
 Summary: Runs a SQL script against a PostgreSQL, SQLite, MariaDB/MySQL, DuckDB, Firebird, MS-Access, MS-SQL-Server, or Oracle database, or an ODBC DSN. Provides metacommands to import and export data, copy data between databases, conditionally execute SQL and metacommands, and dynamically alter SQL and metacommands with substitution variables.
 Project-URL: Homepage, https://execsql2.readthedocs.io
 Project-URL: Repository, https://github.com/geocoug/execsql
@@ -384,7 +384,7 @@ execsql-format --no-sql --in-place scripts/
 ```yaml
 repos:
   - repo: https://github.com/geocoug/execsql
-    rev: v2.22.4
+    rev: v2.22.5
     hooks:
       - id: execsql-format
 ```

{execsql2-2.22.4.dist-info → execsql2-2.22.5.dist-info}/RECORD

@@ -3,7 +3,7 @@ execsql/__main__.py,sha256=HdbK-SAhyUmfB6xINY5AzxdMSxGzWSGEG_2dv42Jn64,315
 execsql/api.py,sha256=EOtBwMnm-8fmB_KmvxyjglR6C7CjdYhCPBisJ9CFfJg,22227
 execsql/config.py,sha256=txKziUJLV7UPonj2sGFgkmlinDdSV50aHmR_cuCbEIo,34369
 execsql/exceptions.py,sha256=GlphAiR8iZu4UR4U50WgSqEUMd9_8AeQJ_hhPcJOeQU,9381
-execsql/format.py,sha256=3w-m0EsHapBdSKx6oTFWYlXspFLdUR_RF79gMvHhrqQ,34440
+execsql/format.py,sha256=tfg0m3EbDVRlAR3ZhwV8BjgMFckL3u3IdldyCONrutU,41400
 execsql/models.py,sha256=-DWTVrPNwM-n1h3e3rXMoBv9-wpWVdZahyy4RB7fRRE,15502
 execsql/parser.py,sha256=aOxZ7Mzgvk7G4-nvy9Bgz7xlaZm6oo81cffDeBk_LCk,15597
 execsql/plugins.py,sha256=l82VWa5NpvOQWD6vcO-WKP-vtn0wp9__wWDFAbwxZg0,12838
@@ -101,23 +101,23 @@ execsql/utils/numeric.py,sha256=_7bi5R6pHjMCkOPne_Sh5nUj-B8e9g6WFmjptFHaunA,1573
 execsql/utils/regex.py,sha256=B3mSBhRRx_DpUnX_t9IoYd05KvWAebiLMCZqQAmEwcw,3885
 execsql/utils/strings.py,sha256=UQNjpRCEFa1UO6feU-M-9e24wWAvizs_iu_4fFusLxo,8516
 execsql/utils/timer.py,sha256=eDYf5VzCNFk7oo90InJucUm3XcBdhYMogjZMqeg9xzc,1899
-execsql2-2.22.4.data/data/execsql2_extras/README.md,sha256=vX4NTL095dUoA_hesyRMGYBorEZ_Y_tJ9qrd-MVV09I,5032
-execsql2-2.22.4.data/data/execsql2_extras/config_settings.sqlite,sha256=aY5cxR7Q7J6zJ4bC9lu5mHUrhy211Cq3MNKPQVCt02E,20480
-execsql2-2.22.4.data/data/execsql2_extras/example_config_prompt.sql,sha256=2e8KzzVWhho8KxYVHETSVmZdhW7wodioDsqBLSL6m4s,7487
-execsql2-2.22.4.data/data/execsql2_extras/make_config_db.sql,sha256=WwyC6dK-Eh5CAVppiBCDHqiI1_wEI9U95Ytpr4lsZkg,8726
-execsql2-2.22.4.data/data/execsql2_extras/md_compare.sql,sha256=qYYVAjSeHZzjszxV3Bv6bg8Ckbq2kMHl87_gh4sywMU,24140
-execsql2-2.22.4.data/data/execsql2_extras/md_glossary.sql,sha256=hkZ2Onn57LAKKsuXxzhR8tPtcWXkmWEQkwPE58-Tm2k,10796
-execsql2-2.22.4.data/data/execsql2_extras/md_upsert.sql,sha256=_CAK4BzEboRXTNy03SJR-oOjcEdSNMuRBPL6noWUptY,112560
-execsql2-2.22.4.data/data/execsql2_extras/pg_compare.sql,sha256=1zJd4hVUKHR0tncc2qTBC9B4qVV4Us2ITkJpsjN3tMw,24352
-execsql2-2.22.4.data/data/execsql2_extras/pg_glossary.sql,sha256=IKuwna-_8b20ljSkXZruuiQigrCpo7ueQdUqd1MXiuI,9908
-execsql2-2.22.4.data/data/execsql2_extras/pg_upsert.sql,sha256=HpPJtTHvpEjQy03j-3iPxDEOHMRkudOg7O4D4YR38UI,108315
-execsql2-2.22.4.data/data/execsql2_extras/script_template.sql,sha256=2J35ddZPguJ-vwTsz83wErv0jiWUyJcdW_JM0mNKDXA,11155
-execsql2-2.22.4.data/data/execsql2_extras/ss_compare.sql,sha256=j1qVNUPXQsEU7-DoVgDJCGcE0EuIl7whLBT3fgeiMAo,24833
-execsql2-2.22.4.data/data/execsql2_extras/ss_glossary.sql,sha256=2gLxv34xzKt0vy7hSzJH7a9JiMC3ETrv9MofxQwAibU,13065
-execsql2-2.22.4.data/data/execsql2_extras/ss_upsert.sql,sha256=G_8rQ0VzuKIZHWs24O_WrfzpC5S27R1JsL-bFBR3SUQ,117730
-execsql2-2.22.4.dist-info/METADATA,sha256=gD2Dm8BheKQrFz6xAs7KyfpqGzs93Rpv7vxWyPWXeQE,20382
-execsql2-2.22.4.dist-info/WHEEL,sha256=zOwg4jB6zX2kU910N-cMawjivD6tO8NEWvE12je1bVk,87
-execsql2-2.22.4.dist-info/entry_points.txt,sha256=sUOxkM-dN1eBGGpSpDLsAaE0yNXYQKWZAfxPOlMkQyk,90
-execsql2-2.22.4.dist-info/licenses/LICENSE.txt,sha256=LBdhuxejF8_bLCHZ2kWfmDXpDGUu914Gbd6_3JjCRe0,676
-execsql2-2.22.4.dist-info/licenses/NOTICE,sha256=McYzgxYav3U1OaVsY4Su1sfBrfmplpRdA9b6-gCDQCg,342
-execsql2-2.22.4.dist-info/RECORD,,
+execsql2-2.22.5.data/data/execsql2_extras/README.md,sha256=vX4NTL095dUoA_hesyRMGYBorEZ_Y_tJ9qrd-MVV09I,5032
+execsql2-2.22.5.data/data/execsql2_extras/config_settings.sqlite,sha256=aY5cxR7Q7J6zJ4bC9lu5mHUrhy211Cq3MNKPQVCt02E,20480
+execsql2-2.22.5.data/data/execsql2_extras/example_config_prompt.sql,sha256=2e8KzzVWhho8KxYVHETSVmZdhW7wodioDsqBLSL6m4s,7487
+execsql2-2.22.5.data/data/execsql2_extras/make_config_db.sql,sha256=WwyC6dK-Eh5CAVppiBCDHqiI1_wEI9U95Ytpr4lsZkg,8726
+execsql2-2.22.5.data/data/execsql2_extras/md_compare.sql,sha256=qYYVAjSeHZzjszxV3Bv6bg8Ckbq2kMHl87_gh4sywMU,24140
+execsql2-2.22.5.data/data/execsql2_extras/md_glossary.sql,sha256=hkZ2Onn57LAKKsuXxzhR8tPtcWXkmWEQkwPE58-Tm2k,10796
+execsql2-2.22.5.data/data/execsql2_extras/md_upsert.sql,sha256=_CAK4BzEboRXTNy03SJR-oOjcEdSNMuRBPL6noWUptY,112560
+execsql2-2.22.5.data/data/execsql2_extras/pg_compare.sql,sha256=1zJd4hVUKHR0tncc2qTBC9B4qVV4Us2ITkJpsjN3tMw,24352
+execsql2-2.22.5.data/data/execsql2_extras/pg_glossary.sql,sha256=IKuwna-_8b20ljSkXZruuiQigrCpo7ueQdUqd1MXiuI,9908
+execsql2-2.22.5.data/data/execsql2_extras/pg_upsert.sql,sha256=HpPJtTHvpEjQy03j-3iPxDEOHMRkudOg7O4D4YR38UI,108315
+execsql2-2.22.5.data/data/execsql2_extras/script_template.sql,sha256=2J35ddZPguJ-vwTsz83wErv0jiWUyJcdW_JM0mNKDXA,11155
+execsql2-2.22.5.data/data/execsql2_extras/ss_compare.sql,sha256=j1qVNUPXQsEU7-DoVgDJCGcE0EuIl7whLBT3fgeiMAo,24833
+execsql2-2.22.5.data/data/execsql2_extras/ss_glossary.sql,sha256=2gLxv34xzKt0vy7hSzJH7a9JiMC3ETrv9MofxQwAibU,13065
+execsql2-2.22.5.data/data/execsql2_extras/ss_upsert.sql,sha256=G_8rQ0VzuKIZHWs24O_WrfzpC5S27R1JsL-bFBR3SUQ,117730
+execsql2-2.22.5.dist-info/METADATA,sha256=OHGAN6pqfqOTBPDSst372rskkKNgeXUofH9K_mfl3sc,20382
+execsql2-2.22.5.dist-info/WHEEL,sha256=zOwg4jB6zX2kU910N-cMawjivD6tO8NEWvE12je1bVk,87
+execsql2-2.22.5.dist-info/entry_points.txt,sha256=sUOxkM-dN1eBGGpSpDLsAaE0yNXYQKWZAfxPOlMkQyk,90
+execsql2-2.22.5.dist-info/licenses/LICENSE.txt,sha256=LBdhuxejF8_bLCHZ2kWfmDXpDGUu914Gbd6_3JjCRe0,676
+execsql2-2.22.5.dist-info/licenses/NOTICE,sha256=McYzgxYav3U1OaVsY4Su1sfBrfmplpRdA9b6-gCDQCg,342
+execsql2-2.22.5.dist-info/RECORD,,