BUG: f2py markinnerspace for multiple quotations (#19419)

Continuation of #15208

* BUG: markinnerspaces does not handle multiple quotations

* TST: test for markinnerspaces in f2py.crackfortran

* TST: tests for markinnerspaces with "

* DOC: markinnerspaces in f2py.crackfortran

* MAINT: more readable r strings in TestMarkinnerspaces

* ENH: give variables names, change 'inside' to bool

* :TEST, MAINT: add tests and changes from review

* MAINT: lint fixes

* BUG: typo

Co-authored-by: bdvd <bdvd001@gmail.com>

2 files changed, 59 insertions, 20 deletions

diff --git a/numpy/f2py/crackfortran.py b/numpy/f2py/crackfortran.py
index 984bd642b..2f491ed7e 100755
--- a/numpy/f2py/crackfortran.py
+++ b/numpy/f2py/crackfortran.py
@@ -1526,27 +1526,40 @@ def removespaces(expr):
 
 
 def markinnerspaces(line):
-    l = ''
-    f = 0
-    cc = '\''
-    cb = ''
+    """
+    The function replace all spaces in the input variable line which are 
+    surrounded with quotation marks, with the triplet "@_@".
+
+    For instance, for the input "a 'b c'" the function returns "a 'b@_@c'"
+
+    Parameters
+    ----------
+    line : str
+
+    Returns
+    -------
+    str
+
+    """  
+    fragment = ''
+    inside = False
+    current_quote = None
+    escaped = ''
     for c in line:
-        if cb == '\\' and c in ['\\', '\'', '"']:
-            l = l + c
-            cb = c
+        if escaped == '\\' and c in ['\\', '\'', '"']:
+            fragment += c
+            escaped = c
             continue
-        if f == 0 and c in ['\'', '"']:
-            cc = c
-        if c == cc:
-            f = f + 1
-        elif c == cc:
-            f = f - 1
-        elif c == ' ' and f == 1:
-            l = l + '@_@'
+        if not inside and c in ['\'', '"']:
+            current_quote = c
+        if c == current_quote:
+            inside = not inside
+        elif c == ' ' and inside:
+            fragment += '@_@'
             continue
-        l = l + c
-        cb = c
-    return l
+        fragment += c
+        escaped = c  # reset to non-backslash
+    return fragment
 
 
 def updatevars(typespec, selector, attrspec, entitydecl):
diff --git a/numpy/f2py/tests/test_crackfortran.py b/numpy/f2py/tests/test_crackfortran.py
index d26917f0c..140f42cbc 100644
--- a/numpy/f2py/tests/test_crackfortran.py
+++ b/numpy/f2py/tests/test_crackfortran.py
@@ -1,8 +1,8 @@
 import numpy as np
-from numpy.testing import assert_array_equal
+from numpy.testing import assert_array_equal, assert_equal
+from numpy.f2py.crackfortran import markinnerspaces
 from . import util
 from numpy.f2py import crackfortran
-import tempfile
 import textwrap
 
 
@@ -37,6 +37,7 @@ class TestNoSpace(util.F2PyTest):
         assert_array_equal(k, w + 1)
         assert self.module.t0(23) == b'2'
 
+
 class TestPublicPrivate():
     def test_defaultPrivate(self, tmp_path):
         f_path = tmp_path / "mod.f90"
@@ -87,6 +88,7 @@ class TestPublicPrivate():
         assert 'private' not in mod['vars']['seta']['attrspec']
         assert 'public' in mod['vars']['seta']['attrspec']
 
+
 class TestExternal(util.F2PyTest):
     # issue gh-17859: add external attribute support
     code = """
@@ -116,6 +118,7 @@ class TestExternal(util.F2PyTest):
         r = self.module.external_as_attribute(incr)
         assert r == 123
 
+
 class TestCrackFortran(util.F2PyTest):
 
     suffix = '.f90'
@@ -139,3 +142,26 @@ class TestCrackFortran(util.F2PyTest):
     def test_gh2848(self):
         r = self.module.gh2848(1, 2)
         assert r == (1, 2)
+
+
+class TestMarkinnerspaces():
+    # issue #14118: markinnerspaces does not handle multiple quotations
+
+    def test_do_not_touch_normal_spaces(self):
+        test_list = ["a ", " a", "a b c", "'abcdefghij'"]
+        for i in test_list:
+            assert_equal(markinnerspaces(i), i)
+
+    def test_one_relevant_space(self):
+        assert_equal(markinnerspaces("a 'b c' \\\' \\\'"), "a 'b@_@c' \\' \\'")
+        assert_equal(markinnerspaces(r'a "b c" \" \"'), r'a "b@_@c" \" \"')
+
+    def test_ignore_inner_quotes(self):
+        assert_equal(markinnerspaces('a \'b c" " d\' e'),
+                     "a 'b@_@c\"@_@\"@_@d' e")
+        assert_equal(markinnerspaces('a "b c\' \' d" e'),
+                     "a \"b@_@c'@_@'@_@d\" e")
+
+    def test_multiple_relevant_spaces(self):
+        assert_equal(markinnerspaces("a 'b c' 'd e'"), "a 'b@_@c' 'd@_@e'")
+        assert_equal(markinnerspaces(r'a "b c" "d e"'), r'a "b@_@c" "d@_@e"')