more comprehensive fileutils testing

tests/modules/lib_testfileutils/lib.cc

@@ -29,6 +29,7 @@ NCTEST_CTYPE_DICTIONARY
     "int nctest_file_exists( const char * );"
     "const char * nctest_ncgetcwd();"
     "const char * nctest_readEntireFileToString( const char * );"
+    "const char * nctest_ncglob( const char * );"
     ;
 }
 
@@ -77,3 +78,23 @@ NCTEST_CTYPES const char * nctest_ncgetcwd()
   return &buf[0];
 }
 
+
+NCTEST_CTYPES const char * nctest_ncglob(const char * pattern)
+{
+  //For testing, we can get away with just using a large static buffer:
+  static char buf[10485760];//10MB, plenty!
+  try {
+    std::ostringstream ss;
+    bool first = true;
+    for (auto&e : NC::ncglob(pattern) ) {
+      if ( !first )
+        ss << "<<@>>";//separator
+      first = false;
+      ss << e;
+    }
+    buf[0] = '\0';
+    std::strncat(buf,ss.str().c_str(),sizeof(buf)-1);
+  } NCCATCH;
+  return &buf[0];
+}
+

tests/pypath/NCTestUtils/common.py

@@ -105,3 +105,9 @@ def _resolve_print_fct(printfct):
         return biprint
     else:
         return printfct
+
+def explicit_unicode_char(c):
+    #32 is space, <32 are control chars, 127 is DEL.
+    return c if 32<=ord(c)<=126 else r'\u{%i}'%ord(c)
+def explicit_unicode_str(s):
+    return ''.join( explicit_unicode_char(c) for c in s)

tests/pypath/NCTestUtils/loadlib.py

@@ -85,7 +85,7 @@ def add( self, fctname, restype, *argtypes ):
     @property
     def functions(self):
         """Get the name of all available functions in this library"""
-        return [f for f,_,_ in self.__fcts]
+        return sorted([f for f,_,_ in self.__fcts])
 
     def dump(self,prefix=''):
         """Print available functions in this library"""
@@ -94,7 +94,7 @@ def dump(self,prefix=''):
                                                 len(self.__fcts)))
         if not self.__fcts:
             print(f"{prefix}  <no functions defined>")
-        for fctname,restype,argtypes in self.__fcts:
+        for fctname,restype,argtypes in sorted(self.__fcts):
             rt = _ctype_2_str(restype)
             a=', '.join([_ctype_2_str(e) for e in argtypes])
             print(f"{prefix}  {rt} {fctname}({a})")

tests/scripts/fileutils.log

@@ -0,0 +1,13 @@
+Library "TestMod_testfileutils" (4 functions):
+  int nctest_file_exists(const char *)
+  const char * nctest_ncgetcwd()
+  const char * nctest_ncglob(const char *)
+  const char * nctest_readEntireFileToString(const char *)
+Testing ncgetcwd...
+...ncgetcwd ok
+READ> 'some'
+READ> 'multi'
+READ> 'line text'
+ncglob('unicode*/*.ncmat') -> ['unicodedir_test\\u{17664}abc/b\\u{955}.ncmat']
+ncglob('unicode*/*.txt') -> ['unicodedir_test\\u{17664}abc/a.txt', 'unicodedir_test\\u{17664}abc/b.txt']
+ncglob('*abc/*\\u{955}*') -> ['unicodedir_test\\u{17664}abc/b\\u{955}.ncmat']

tests/scripts/fileutils.py

@@ -22,33 +22,73 @@
 
 import NCTestUtils.enable_fpe
 from NCTestUtils.loadlib import Lib
+from NCTestUtils.common import explicit_unicode_str
+import pathlib
+
 lib = Lib('testfileutils')
 lib.dump()
 assert hasattr(lib,'nctest_file_exists')
 assert hasattr(lib,'nctest_ncgetcwd')
+assert hasattr(lib,'nctest_readEntireFileToString')
+assert hasattr(lib,'nctest_ncglob')
+
+_raw_ncglob = lib.nctest_ncglob
+def nctest_ncglob( pattern ):
+    return _raw_ncglob(pattern).split('<<@>>')
+lib.nctest_ncglob = nctest_ncglob
+
+def test1():
+    d = pathlib.Path('.')
+    (d / 'foo.txt').write_text('foo')
+
+    assert lib.nctest_file_exists('./foo.txt')
+    assert lib.nctest_file_exists('foo.txt')
+    assert not lib.nctest_file_exists('bar.txt')
+    assert not lib.nctest_file_exists('./bar.txt')
+    assert not lib.nctest_file_exists('/some/file/that/does/not/exist')
+    print("Testing ncgetcwd...")
+    _cwd = lib.nctest_ncgetcwd()
+    d_abs = d.resolve().absolute()
+    assert d.samefile(d_abs)
+    assert d_abs.samefile ( pathlib.Path(_cwd).resolve().absolute() )
+    assert d_abs.samefile ( _cwd )
+    assert d.samefile ( _cwd )
+    print("...ncgetcwd ok")
+
+def test2():
+    dirname = 'unicodedir_test\u4500abc'
+    subdir = pathlib.Path('.') / dirname
+    subdir.mkdir()
+    for fn in [ 'a.txt', 'b.txt', 'b\u03bb.ncmat' ]:
+        (subdir / fn).write_text('dummy')
+    testtext = 'some\nmulti\nline text\n'
+    (subdir / 'b\u03bb.ncmat').write_text(testtext)
+
+    content = lib.nctest_readEntireFileToString(
+        'unicodedir_test\u4500abc/b\u03bb.ncmat'
+    )
+    assert testtext == content
+    for e in content.splitlines():
+        print('READ>',repr(e))
+
+    def testglob(pattern,nexpect):
+        g = lib.nctest_ncglob(pattern)
+        assert len(g) == nexpect
+        g = list(explicit_unicode_str(e) for e in g)
+        print(f'ncglob({repr(explicit_unicode_str(pattern))}) -> {repr(g)}')
+    testglob('unicode*/*.ncmat',1)
+    testglob('unicode*/*.txt',2)
+    testglob('*abc/*\u03bb*',1)
+
+def main():
+
+
+    test1()
+    test2()
+
+if __name__=='__main__':
+    main()
 
-import pathlib
-d = pathlib.Path('.')
-(d / 'foo.txt').write_text('foo')
-
-assert lib.nctest_file_exists('./foo.txt')
-assert lib.nctest_file_exists('foo.txt')
-assert not lib.nctest_file_exists('bar.txt')
-assert not lib.nctest_file_exists('./bar.txt')
-assert not lib.nctest_file_exists('/some/file/that/does/not/exist')
-print("Testing that two files are identical:")
-_cwd = lib.nctest_ncgetcwd()
-d_abs = d.resolve().absolute()
-assert d.samefile(d_abs)
-print("  1) %s"%d_abs)
-print("  2) %s"%_cwd)
-assert d_abs.samefile ( pathlib.Path(_cwd).resolve().absolute() )
-assert d_abs.samefile ( _cwd )
-assert d.samefile ( _cwd )
-
-#FIXME : Step into dir with on-ascii name and check ncgetcwd ??
-
-#print('cwd',lib.nctest_ncgetcwd())
 
 #FIXME: Much more, including globbing, is_absolute_path("c:[\]bla.ncmat) and
 #whatever is likely to cause issues on Windows.

tests/scripts/vdosgn.py

@@ -65,6 +65,7 @@ def main():
 
     PWLinDistMoments.unit_test(mp)
 
+    #Fixme: add many more here, so we have good coverage.
     for filename in ('Au_sg225',
                      'Polyethylene_CH2'):
         for temp in (10,600):