Merge pull request #2489 from advikkabra/isspace-fix

Fix isspace edge case

Author: anutosh491

integration_tests/test_str_attributes.py

@@ -21,13 +21,15 @@ def lower():
     assert "DDd12Vv" .lower() == "ddd12vv"
     assert "".lower() == ""
 
+
 def upper():
     s: str
     s = "AaaaAABBbbbbBB!@12223BN"
     assert s.upper() == "AAAAAABBBBBBBB!@12223BN"
     assert "DDd12Vv".upper() == "DDD12VV"
     assert "".upper() == ""
 
+
 def strip():
     s: str
     s = "     AASAsaSas    "
@@ -88,14 +90,15 @@ def startswith():
     assert s.startswith("sdd") == False
     assert "".startswith("ok") == False
 
+
 def endswith():
 
     # The following test suite fulfils the control flow graph coverage
     # in terms of Statement Coverage and Branch Coverage associated with endwith() functionality.
 
-    # Case 1: When string is constant and suffix is also constant 
+    # Case 1: When string is constant and suffix is also constant
     assert "".endswith("") == True
-    assert "".endswith(" ") ==  False
+    assert "".endswith(" ") == False
     assert "".endswith("%") == False
     assert "".endswith("a1234PT#$") == False
     assert "".endswith("blah blah") == False
@@ -105,13 +108,12 @@ def endswith():
     assert " rendezvous 5:30 ".endswith("apple") == False
     assert "two plus".endswith("longer than string") == False
 
-
     # Case 2: When string is constant and suffix is variable
     suffix: str
     suffix = ""
     assert "".endswith(suffix) == True
     suffix = " "
-    assert "".endswith(suffix) ==  False
+    assert "".endswith(suffix) == False
     suffix = "5:30 "
     assert " rendezvous 5:30 ".endswith(suffix) == True
     suffix = ""
@@ -138,13 +140,14 @@ def endswith():
     suffix = "apple"
     assert s.endswith(suffix) == False
 
+
 def partition():
-    
-    # Note: Both string or seperator cannot be empty 
-    # Case 1: When string is constant and seperator is also constant 
-    assert "   ".partition(" ") == (""," ","  ")
-    assert "apple mango".partition(" ") == ("apple"," ","mango")
-    assert "applemango".partition("afdnjkfsn") ==  ("applemango","","")
+
+    # Note: Both string or seperator cannot be empty
+    # Case 1: When string is constant and seperator is also constant
+    assert "   ".partition(" ") == ("", " ", "  ")
+    assert "apple mango".partition(" ") == ("apple", " ", "mango")
+    assert "applemango".partition("afdnjkfsn") == ("applemango", "", "")
     assert "applemango".partition("an") == ("applem", "an", "go")
     assert "applemango".partition("mango") == ("apple", "mango", "")
     assert "applemango".partition("applemango") == ("", "applemango", "")
@@ -154,15 +157,17 @@ def partition():
     # Case 2: When string is constant and seperator is variable
     seperator: str
     seperator = " "
-    assert "   ".partition(seperator) == (""," ","  ")
+    assert "   ".partition(seperator) == ("", " ", "  ")
     seperator = " "
-    assert "apple mango".partition(seperator) ==  ("apple"," ","mango")
+    assert "apple mango".partition(seperator) == ("apple", " ", "mango")
     seperator = "5:30 "
-    assert " rendezvous 5:30 ".partition(seperator) == (" rendezvous ", "5:30 ", "")
+    assert " rendezvous 5:30 ".partition(
+        seperator) == (" rendezvous ", "5:30 ", "")
     seperator = "^&"
     assert "@#$%^&*()#!".partition(seperator) == ("@#$%", "^&", "*()#!")
     seperator = "daddada "
-    assert " rendezvous 5:30 ".partition(seperator) == (" rendezvous 5:30 ", "", "")
+    assert " rendezvous 5:30 ".partition(
+        seperator) == (" rendezvous 5:30 ", "", "")
     seperator = "longer than string"
     assert "two plus".partition(seperator) == ("two plus", "", "")
 
@@ -182,6 +187,7 @@ def partition():
     seperator = "apple"
     assert s.partition(seperator) == ("rendezvous 5", "", "")
 
+
 def is_lower():
     # Case 1: When constant string is present
     assert "".islower() == False
@@ -204,8 +210,9 @@ def is_lower():
     s = "apple is a fruit"
     assert s.islower() == True
 
+
 def is_upper():
-        # Case 1: When constant string is present
+    # Case 1: When constant string is present
     assert "".isupper() == False
     assert "apple".isupper() == False
     assert "4432632479".isupper() == False
@@ -226,6 +233,7 @@ def is_upper():
     s = "APPLE IS A FRUIT"
     assert s.isupper() == True
 
+
 def is_decimal():
     # Case 1: When constant string is present
     assert "".isdecimal() == False
@@ -251,6 +259,7 @@ def is_decimal():
     s = "12 34"
     assert s.isdecimal() == False
 
+
 def is_ascii():
     # Case 1: When constant string is present
     assert "".isascii() == True
@@ -280,12 +289,16 @@ def is_ascii():
 def is_space():
     assert "\n".isspace() == True
     assert "    ".isspace() == True
-    assert "\r".isspace() == True 
+    assert "\r".isspace() == True
+    assert "".isspace() == False
 
-    s:str = " "
-    assert s.isspace() == True   
+    s: str = " "
+    assert s.isspace() == True
     s = "a"
     assert s.isspace() == False
+    s = ""
+    assert s.isspace() == False
+
 
 def check():
     capitalize()
@@ -303,4 +316,5 @@ def check():
     is_ascii()
     is_space()
 
+
 check()

src/lpython/semantics/python_ast_to_asr.cpp

@@ -7107,18 +7107,18 @@ class BodyVisitor : public CommonVisitor<BodyVisitor> {
                     Return True if all cased characters in the string are uppercase and there is at least one cased character, False otherwise.
                 */
                 bool is_cased_present = false;
-                bool is_lower = true;
+                bool is_upper = true;
                 for (auto &i : s_var) {
                     if ((i >= 'A' && i <= 'Z') || (i >= 'a' && i <= 'z')) {
                         is_cased_present = true;
                         if(!(i >= 'A' && i <= 'Z')) {
-                            is_lower = false;
+                            is_upper = false;
                             break;
                         }
                     }
                 }
-                is_lower = is_lower && is_cased_present;
-                tmp = ASR::make_LogicalConstant_t(al, loc, is_lower,
+                is_upper = is_upper && is_cased_present;
+                tmp = ASR::make_LogicalConstant_t(al, loc, is_upper,
                         ASRUtils::TYPE(ASR::make_Logical_t(al, loc, 4)));
                 return;
             } else if(attr_name == "isdecimal") {
@@ -7160,13 +7160,13 @@ characters, as defined by CPython. Return false otherwise. For now we use the
 std::isspace function, but if we later discover that it differs from CPython,
 we will have to use something else.
                 */
-                bool is_space = true;
-                    for (char i : s_var) {
-                        if (!std::isspace(static_cast<unsigned char>(i))) {
-                            is_space = false;
-                            break;
-                        }
+                bool is_space = (s_var.size() != 0);
+                for (char i : s_var) {
+                    if (!std::isspace(static_cast<unsigned char>(i))) {
+                        is_space = false;
+                        break;
                     }
+                }
                 tmp = ASR::make_LogicalConstant_t(al, loc, is_space,
                         ASRUtils::TYPE(ASR::make_Logical_t(al, loc, 4)));
                 return;

src/runtime/lpython_builtin.py

@@ -955,7 +955,7 @@ def _lpython_str_isdecimal(s: str) -> bool:
 
 @overload
 def _lpython_str_isascii(s: str) -> bool:
-    if(len(s) == 0):
+    if len(s) == 0:
         return True
     i: str
     for i in s:
@@ -964,6 +964,8 @@ def _lpython_str_isascii(s: str) -> bool:
     return True
 
 def _lpython_str_isspace(s:str) -> bool:
+    if len(s) == 0:
+        return False
     ch: str 
     for ch in s:
         if ch != ' ' and ch != '\t' and ch != '\n' and ch != '\r' and ch != '\f' and ch != '\v':