oaknut-basic 12.9.0__tar.gz → 12.10.1__tar.gz

{oaknut_basic-12.9.0/src/oaknut_basic.egg-info → oaknut_basic-12.10.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: oaknut-basic
-Version: 12.9.0
+Version: 12.10.1
 Summary: BBC BASIC tools: program tokeniser/de-tokeniser and PRINT#/INPUT# data-file reader/writer
 Author-email: Robert Smallshire <robert@smallshire.org.uk>
 License-Expression: MIT

{oaknut_basic-12.9.0 → oaknut_basic-12.10.1}/src/oaknut/basic/__init__.py

@@ -79,7 +79,7 @@ from oaknut.basic.tokens import (
     TOKEN_TO_KEYWORD,
 )
 
-__version__ = "12.9.0"
+__version__ = "12.10.1"
 
 # Canonical load addresses for BBC BASIC programs on each host.
 # Programs saved by *SAVE on a real machine use these by default.

{oaknut_basic-12.9.0 → oaknut_basic-12.10.1}/src/oaknut/basic/numbering.py

@@ -14,9 +14,35 @@ stream plumbing on top.
 
 from __future__ import annotations
 
+import re
+
 DEFAULT_LINE_NUMBER = 10
 DEFAULT_LINE_STEP = 10
 
+# A BBC BASIC line ends only at a carriage return (or a host newline);
+# Python's str.splitlines() additionally breaks on &0B, &0C, &1C-&1E and
+# others, which occur legitimately inside string literals (mode-7 / VDU
+# control codes), so split strictly on the real terminators only. This
+# definition is shared with the tokeniser, which imports it from here.
+LINE_SEPARATOR_RE = re.compile(r"\r\n|\r|\n")
+
+
+def split_source_lines(source: str) -> list[str]:
+    """Split BBC BASIC source into lines on real terminators only.
+
+    Matches :meth:`str.splitlines` semantics for the terminators a BBC
+    BASIC line actually ends on (``"\\n"``, ``"\\r"``, ``"\\r\\n"``) — an
+    empty *source* yields no lines and a single trailing terminator is
+    not treated as introducing a final empty line — but, unlike
+    :meth:`str.splitlines`, never breaks on the in-string control codes
+    (&0B, &0C, &1C-&1E, ...) that BBC BASIC string literals legitimately
+    contain.
+    """
+    lines = LINE_SEPARATOR_RE.split(source)
+    if lines and lines[-1] == "":
+        lines.pop()
+    return lines
+
 
 def number_lines(
     source: str,
@@ -54,7 +80,7 @@ def number_lines(
 
     number = start
     numbered = []
-    for line in source.splitlines():
+    for line in split_source_lines(source):
         numbered.append(f"{number} {line}")
         number += step
 

{oaknut_basic-12.9.0 → oaknut_basic-12.10.1}/src/oaknut/basic/tokeniser.py

@@ -27,8 +27,6 @@ the CLI does it at its I/O boundary.
 
 from __future__ import annotations
 
-import re
-
 from oaknut.basic.exceptions import (
     AlreadyNumberedError,
     LineNumberOrderError,
@@ -37,7 +35,12 @@ from oaknut.basic.exceptions import (
     UnnumberedLineError,
 )
 from oaknut.basic.linenumber import MAX_LINE_NUMBER, encode_line_number
-from oaknut.basic.numbering import DEFAULT_LINE_NUMBER, DEFAULT_LINE_STEP, number_lines
+from oaknut.basic.numbering import (
+    DEFAULT_LINE_NUMBER,
+    DEFAULT_LINE_STEP,
+    LINE_SEPARATOR_RE,
+    number_lines,
+)
 from oaknut.basic.tokens import (
     FLAG_CONDITIONAL,
     FLAG_FN_PROC,
@@ -56,15 +59,9 @@ from oaknut.basic.tokens import (
 _CR = 0x0D
 _END_MARKER = 0xFF
 
-# A BBC BASIC line ends only at a carriage return (or a host newline);
-# Python's str.splitlines() additionally breaks on &0B, &0C, &1C-&1E and
-# others, which occur legitimately inside string literals (mode-7 / VDU
-# control codes), so split strictly on the real terminators only.
-_LINE_SEPARATOR_RE = re.compile(r"\r\n|\r|\n")
-
 
 def _split_source_lines(source: str) -> list[str]:
-    return _LINE_SEPARATOR_RE.split(source)
+    return LINE_SEPARATOR_RE.split(source)
 
 # Keyword entries grouped by first character, preserving ROM order within
 # each group, so the crunch only scans the relevant group.

{oaknut_basic-12.9.0 → oaknut_basic-12.10.1/src/oaknut_basic.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: oaknut-basic
-Version: 12.9.0
+Version: 12.10.1
 Summary: BBC BASIC tools: program tokeniser/de-tokeniser and PRINT#/INPUT# data-file reader/writer
 Author-email: Robert Smallshire <robert@smallshire.org.uk>
 License-Expression: MIT

{oaknut_basic-12.9.0 → oaknut_basic-12.10.1}/tests/test_numbering.py

@@ -47,6 +47,13 @@ class TestNumberLines:
     def test_blank_lines_are_numbered_too(self):
         assert basic.number_lines("A\n\nB") == "10 A\n20 \n30 B"
 
+    def test_control_codes_in_string_literal_do_not_split_the_line(self):
+        # VDU / mode-7 control codes (here &0C, &1C) appear legitimately
+        # inside string literals; str.splitlines() would break on them,
+        # corrupting a single program line into several numbered lines.
+        source = 'PRINT "' + chr(0x0C) + chr(0x1C) + 'banner"'
+        assert basic.number_lines(source) == '10 PRINT "\x0c\x1cbanner"'
+
     def test_internal_references_are_left_untouched(self):
         # The facade only prepends numbers; it does not rewrite GOTO etc.
         source = "GOTO 20\nEND"