Refactor string attributes and add constant string implementations (lcompilers#2524)

Author: advikkabra

integration_tests/test_str_01.py

@@ -37,9 +37,6 @@ def test_str_slice():
     # TODO:
     # assert a[0:5:-1] == ""
 
-<<<<<<< HEAD
-
-=======
 def test_str_isalpha():
     a: str = "helloworld"
     b: str = "hj kl"
@@ -56,7 +53,6 @@ def test_str_isalpha():
     assert res3 == False
     assert res4 == False
     assert res5 == False
->>>>>>> 08e1f4502 (Fixes empty string bug in str isalpha method)
 
 
 def test_str_title():
@@ -179,8 +175,6 @@ def check():
     test_str_join_empty_list()
     test_constant_str_subscript()
     test_str_title()
-    test_str_istitle()
-    test_str_isalpha()
     test_str_split()
 
 check()

integration_tests/test_str_attributes.py

@@ -343,6 +343,53 @@ def is_alpha():
     s = " "
     assert s.isalpha() == False
 
+def is_alpha():
+    a: str = "helloworld"
+    b: str = "hj kl"
+    c: str = "a12(){}A"
+    d: str = " "
+    e: str = ""
+    res: bool = a.isalpha()
+    res2: bool = b.isalpha()
+    res3: bool = c.isalpha()
+    res4: bool = d.isalpha()
+    res5: bool = e.isalpha()
+    assert res == True
+    assert res2 == False
+    assert res3 == False
+    assert res4 == False
+    assert res5 == False
+
+    assert "helloworld".isalpha() == True
+    assert "hj kl".isalpha() == False
+    assert "a12(){}A".isalpha() == False
+    assert " ".isalpha() == False
+    assert "".isalpha() == False
+
+
+def is_title():
+    a: str = "Hello World"
+    b: str = "Hj'kl"
+    c: str = "hELlo wOrlD"
+    d: str = " Hello"
+    e: str = " "
+    res: bool = a.istitle()
+    res2: bool = b.istitle()
+    res3: bool = c.istitle()
+    res4: bool = d.istitle()
+    res5: bool = e.istitle()
+    assert res == True
+    assert res2 == False 
+    assert res3 == False
+    assert res4 == True
+    assert res5 == False
+
+    assert "Hello World".istitle() == True
+    assert "Hj'kl".istitle() == False
+    assert "hELlo wOrlD".istitle() == False
+    assert " Hello".istitle() == True
+    assert " ".istitle() == False
+
 def is_space():
     assert "\n".isspace() == True
     assert "    ".isspace() == True
@@ -357,6 +404,7 @@ def is_space():
     assert s.isspace() == False
 
 
+
 def check():
     capitalize()
     lower()
@@ -372,6 +420,8 @@ def check():
     is_upper()
     is_decimal()
     is_ascii()
+    is_alpha()
+    is_title()
     is_space()
     is_title()
     is_alpha()

src/lpython/semantics/python_ast_to_asr.cpp

@@ -6577,26 +6577,6 @@ class BodyVisitor : public CommonVisitor<BodyVisitor> {
             arg.loc = loc;
             arg.m_value = s_var;
             fn_args.push_back(al, arg);
-        } else if (attr_name == "isalpha") {
-            if (args.size() != 0) {
-                throw SemanticError("str.isalpha() takes no arguments",
-                    loc);
-            }
-            fn_call_name = "_lpython_str_isalpha";
-            ASR::call_arg_t arg;
-            arg.loc = loc;
-            arg.m_value = s_var;
-            fn_args.push_back(al, arg);
-        } else if (attr_name == "istitle") {
-            if (args.size() != 0) {
-                throw SemanticError("str.istitle() takes no arguments",
-                    loc);
-            }
-            fn_call_name = "_lpython_str_istitle";
-            ASR::call_arg_t arg;
-            arg.loc = loc;
-            arg.m_value = s_var;
-            fn_args.push_back(al, arg);
         } else if (attr_name == "title") {
             if (args.size() != 0) {
                 throw SemanticError("str.title() takes no arguments",
@@ -7116,7 +7096,7 @@ class BodyVisitor : public CommonVisitor<BodyVisitor> {
                 * islower() method is limited to English Alphabets currently
                 * TODO: We can support other characters from Unicode Library
             */
-            std::vector<std::string> validation_methods{"lower", "upper", "decimal", "ascii", "space","alpha","title"};  // Database of validation methods supported
+            std::vector<std::string> validation_methods{"lower", "upper", "decimal", "ascii", "space", "alpha", "title"};  // Database of validation methods supported
             std::string method_name = attr_name.substr(2);
             if(std::find(validation_methods.begin(),validation_methods.end(), method_name) == validation_methods.end()) {
                 throw SemanticError("String method not implemented: " + attr_name, loc);
@@ -7272,6 +7252,57 @@ we will have to use something else.
                 tmp = ASR::make_LogicalConstant_t(al, loc, is_space,
                         ASRUtils::TYPE(ASR::make_Logical_t(al, loc, 4)));
                 return;
+            } else if (attr_name == "isalpha") {
+                /*
+                    * Specification -
+                    Return True if all characters in the string are alphabets, 
+                    and there is at least one character in the string.
+                */
+                bool is_alpha = (s_var.size() != 0);
+                for (auto &i : s_var) {
+                    if (!((i >= 'A' && i <= 'Z') || (i >= 'a' && i <= 'z'))) {
+                        is_alpha = false;
+                        break;
+                    }
+                }
+                tmp = ASR::make_LogicalConstant_t(al, loc, is_alpha,
+                        ASRUtils::TYPE(ASR::make_Logical_t(al, loc, 4)));
+                return;
+            } else if (attr_name == "istitle") {
+                /*
+                    * Specification -
+                    Returns True if all words in the string are in title case, 
+                    and there is at least one character in the string.
+                */
+                bool is_title = (s_var.size() != 0);
+
+                bool in_word = false; // Represents if we are in a word or not
+                bool is_alpha_present = false;
+                for (auto &i : s_var) {
+                    if (i >= 'A' && i <= 'Z') {
+                        is_alpha_present = true;
+                        if (in_word) {
+                            // We have come across an uppercase character in the middle of a word
+                            is_title = false;
+                            break;
+                        } else {
+                            in_word = true;
+                        }
+                    } else if (i >= 'a' && i <= 'z') {
+                        is_alpha_present = true;
+                        if (!in_word) {
+                            //We have come across a lowercase character at the start of a word
+                            is_title = false;
+                            break;
+                        }
+                    } else {
+                        in_word = false;
+                    }
+                }
+                is_title = is_title && is_alpha_present;
+                tmp = ASR::make_LogicalConstant_t(al, loc, is_title,
+                        ASRUtils::TYPE(ASR::make_Logical_t(al, loc, 4)));
+                return;
             } else {
                 throw SemanticError("'str' object has no attribute '" + attr_name + "'", loc);
             }

src/runtime/lpython_builtin.py

@@ -764,9 +764,7 @@ def _lpython_str_istitle(s: str) -> bool:
     ch: str
     only_whitespace: bool = True 
     for ch in s:
-        if (ch == ' ' or ch == '\t' or ch == '\n') and word_start:
-            continue  # Found a space character at the start of a word
-        elif ch.isalpha() and (ord('A') <= ord(ch) and ord(ch) <= ord('Z')):
+        if ch.isalpha() and (ord('A') <= ord(ch) and ord(ch) <= ord('Z')):
             only_whitespace = False
             if word_start:
                 word_start = False