sql-blocks 1.25.610999999999__py3-none-any.whl → 1.25.6139999999999__py3-none-any.whl

sql_blocks/sql_blocks.py

@@ -81,9 +81,23 @@ class SQLObject:
  
     @staticmethod
     def get_separator(key: str) -> str:
-        appendix = {WHERE: r'\s+and\s+|', FROM: r'\s+join\s+|\s+JOIN\s+'}
+        if key == WHERE:
+            return r'\s+and\s+|\s+AND\s+'
+        appendix = {FROM: r'\s+join\s+|\s+JOIN\s+'}
         return KEYWORD[key][0].format(appendix.get(key, ''))
 
+    @staticmethod
+    def contains_CASE_statement(text: str) -> bool:
+        return re.search(r'\bCASE\b', text, re.IGNORECASE)
+
+    @classmethod
+    def split_fields(cls, text: str, key: str) -> list:
+        if key == SELECT and cls.contains_CASE_statement(text):
+            return Case.parse(text)
+        text = re.sub(r'\s+', ' ', text)
+        separator = cls.get_separator(key)
+        return re.split(separator, text)
+
     @staticmethod
     def is_named_field(fld: str, name: str='') -> bool:
         return re.search(fr'(\s+as\s+|\s+AS\s+){name}', fld)
@@ -103,16 +117,13 @@ class SQLObject:
                 result += re.split(r'([=()]|<>|\s+ON\s+|\s+on\s+)', fld)
             return result
         def cleanup(text: str) -> str:
-            if re.search(r'^CASE\b', text):
+            # if re.search(r'^CASE\b', text):
+            if self.contains_CASE_statement(text):
                 return text
             text = re.sub(r'[\n\t]', ' ', text)
             if exact:
                 text = text.lower()
             return text.strip()
-        def split_fields(text: str) -> list:
-            if key == SELECT:
-                return Case.parse(text)
-            return re.split(separator, text)
         def field_set(source: list) -> set:
             return set(
                 (
@@ -122,14 +133,13 @@ class SQLObject:
                     re.sub(pattern, '', cleanup(fld))
                 )
                 for string in disassemble(source)
-                for fld in split_fields(string)
+                for fld in self.split_fields(string, key)
             )       
         pattern = KEYWORD[key][1] 
         if exact:
             if key == WHERE:
                 pattern = r'["\']| '
             pattern += f'|{PATTERN_PREFIX}'
-        separator = self.get_separator(key)
         s1 = field_set(search_list)
         s2 = field_set(self.values.get(key, []))
         if exact:
@@ -727,7 +737,7 @@ class Case:
                 result += block.fields
                 block.fields = []
             elif word not in RESERVED_WORDS:
-                result.append(word)
+                result.append(word.replace(',', ''))
             last_word = word
         return result
 
@@ -1213,6 +1223,10 @@ class PandasLanguage(DataAnalysisLanguage):
     pattern = '{_from}{where}{select}{group_by}{order_by}'
     has_default = {key: False for key in KEYWORD}
     file_extension = FileExtension.CSV
+    HEADER_IMPORT_LIB  = ['import pandas as pd']
+    LIB_INITIALIZATION = ''
+    FIELD_LIST_FMT = '[[{}{}]]'
+    PREFIX_LIBRARY = 'pd.'
 
     def add_field(self, values: list) -> str:
         def line_field_fmt(field: str) -> str:
@@ -1221,30 +1235,43 @@ class PandasLanguage(DataAnalysisLanguage):
             )
         common_fields = self.split_agg_fields(values)
         if common_fields:
-            return '[[{}\n]]'.format(
-                ','.join(line_field_fmt(fld) for fld in common_fields)
+            return self.FIELD_LIST_FMT.format(
+                ','.join(line_field_fmt(fld) for fld in common_fields),
+                self.LINE_BREAK
             )
         return ''
 
+    def merge_tables(self, elements: list, main_table: str) -> str:
+        a1, f1, a2, f2 = elements
+        return "\n\ndf_{} = pd.merge(\n\tdf_{}, df_{}, left_on='{}', right_on='{}', how='{}'\n)\n".format(
+            main_table, self.names[a1], self.names[a2], f1, f2, 'inner'
+        )
+
     def get_tables(self, values: list) -> str:
-        result = 'import pandas as pd'
-        names = {}
+        result = '\n'.join(self.HEADER_IMPORT_LIB) + '\n'
+        if self.LIB_INITIALIZATION:
+            result += f'\n{self.LIB_INITIALIZATION}'
+        self.names = {}
         for table in values:
             table, *join = [t.strip() for t in re.split('JOIN|LEFT|RIGHT|ON', table) if t.strip()]
             alias, table = SQLObject.split_alias(table)
-            result += "\ndf_{table} = pd.{func}('{table}.{ext}')".format(
-                table=table, func=self.file_extension.value, ext=self.file_extension.name.lower()
+            result += "\ndf_{table} = {prefix}{func}('{table}.{ext}')".format(
+                prefix=self.PREFIX_LIBRARY, func=self.file_extension.value,
+                table=table, ext=self.file_extension.name.lower()
             )
-            names[alias] = table
+            self.names[alias] = table
             if join:
-                a1, f1, a2, f2 = [r.strip() for r in re.split('[().=]', join[-1]) if r]
-                result += "\n\ndf_{} = pd.merge(\n\tdf_{}, df_{}, left_on='{}', right_on='{}', how='{}'\n)\n".format(
-                    last_table, names[a1], names[a2], f1, f2, 'inner'
-                )
+                result += self.merge_tables([
+                    r.strip() for r in re.split('[().=]', join[-1]) if r
+                ], last_table)
             last_table = table
         _, table = SQLObject.split_alias(values[0])
-        result += f'\ndf = df_{table}\n\ndf = df\n'
+        result += f'\ndf = df_{table}\n\ndf = df'
         return result
+    
+    def split_condition_elements(self, expr: str) -> list:
+        expr = self.remove_alias(expr)
+        return [t for t in re.split(r'(\w+)', expr) if t.strip()]
 
     def extract_conditions(self, values: list) -> str:
         conditions = []
@@ -1254,8 +1281,7 @@ class PandasLanguage(DataAnalysisLanguage):
             3: '.str.contains(',
         }
         for expr in values:
-            expr = self.remove_alias(expr)
-            field, op, *const = [t for t in re.split(r'(\w+)', expr) if t.strip()]
+            field, op, *const = self.split_condition_elements(expr)
             if op.upper() == 'LIKE' and len(const) == 3:
                 level = 0
                 if '%' in const[0]:
@@ -1312,6 +1338,73 @@ class PandasLanguage(DataAnalysisLanguage):
         return ''
 
 
+class SparkLanguage(PandasLanguage):
+    HEADER_IMPORT_LIB = [
+        'from pyspark.sql import SparkSession',
+        'from pyspark.sql.functions import col, avg, sum, count'
+    ]
+    FIELD_LIST_FMT = '.select({}{})'
+    PREFIX_LIBRARY = 'pyspark.pandas.'
+
+    def merge_tables(self, elements: list, main_table: str) -> str:
+        a1, f1, a2, f2 = elements
+        COMMAND_FMT = """{cr}
+        df_{result} = df_{table1}.join(
+            {indent}df_{table2},
+            {indent}df_{table1}.{fk_field}{op}df_{table2}.{primary_key}{cr}
+        )
+        """
+        return re.sub(r'\s+', '', COMMAND_FMT).format(
+            result=main_table, cr=self.LINE_BREAK, indent=self.TABULATION,
+            table1=self.names[a1], table2=self.names[a2],
+            fk_field=f1, primary_key=f2, op=' == '
+        )
+
+    def extract_conditions(self, values: list) -> str:
+        conditions = []
+        for expr in values:
+            field, op, *const = self.split_condition_elements(expr)
+            const = ''.join(const)
+            if op.upper() == 'LIKE':
+                line = f"\n\t( col('{field}').like({const}) )"
+            else:
+                line = f"\n\t( col('{field}') {op} {const} )"
+            conditions.append(line)
+        if not conditions:
+            return ''
+        return '.filter({}\n)'.format(
+            '\n\t&'.join(conditions)
+        )
+
+    def sort_by(self, values: list) -> str:
+        if not values:
+            return ''
+        return '.orderBy({}{}{})'.format(
+            self.TABULATION,
+            self.clean_values(values),
+            self.LINE_BREAK
+        )
+
+    def set_group(self, values: list) -> str:
+        result = '.groupBy({}{}{})'.format(
+            self.TABULATION,
+            self.clean_values(values),
+            self.LINE_BREAK
+        )
+        if self.aggregation_fields:            
+            result += '.agg('
+            for field in self.aggregation_fields:
+                func, field, *alias = re.split(r'[()]|\s+as\s+|\s+AS\s+', field)
+                result += "{}{}('{}')".format(
+                    self.TABULATION, func.lower(), 
+                    field if field else '*'
+                )
+                if alias:
+                    result += f".alias('{alias[-1]}')"
+            result += '\n)'
+        return result
+
+
 class Parser:
     REGEX = {}
 
@@ -1359,7 +1452,7 @@ class SQLParser(Parser):
     def prepare(self):
         keywords = '|'.join(k + r'\b' for k in KEYWORD)
         flags = re.IGNORECASE + re.MULTILINE
-        self.REGEX['keywords'] = re.compile(f'({keywords}|[*])', flags)
+        self.REGEX['keywords'] = re.compile(f'({keywords})', flags)
         self.REGEX['subquery'] = re.compile(r'(\w\.)*\w+ +in +\(SELECT.*?\)', flags)
 
     def eval(self, txt: str):
@@ -1417,13 +1510,12 @@ class SQLParser(Parser):
                 for key in USUAL_KEYS:
                     if not key in values:
                         continue
-                    separator = self.class_type.get_separator(key)
                     cls = {
                         ORDER_BY: OrderBy, GROUP_BY: GroupBy
                     }.get(key, Field)
                     obj.values[key] = [
                         cls.format(fld, obj)
-                        for fld in re.split(separator, values[key])
+                        for fld in self.class_type.split_fields(values[key], key)
                         if (fld != '*' and len(tables) == 1) or obj.match(fld, key)
                     ]
                 result[obj.alias] = obj
@@ -2110,5 +2202,3 @@ def detect(text: str, join_queries: bool = True, format: str='') -> Select | lis
         result += query
     return result
 # ===========================================================================================//
-
-

{sql_blocks-1.25.610999999999.dist-info → sql_blocks-1.25.6139999999999.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sql_blocks
-Version: 1.25.610999999999
+Version: 1.25.6139999999999
 Summary: Allows you to create objects for parts of SQL query commands. Also to combine these objects by joining them, adding or removing parts...
 Home-page: https://github.com/julio-cascalles/sql_blocks
 Author: Júlio Cascalles

sql_blocks-1.25.6139999999999.dist-info/RECORD

@@ -0,0 +1,7 @@
+sql_blocks/__init__.py,sha256=5ItzGCyqqa6kwY8wvF9kapyHsAiWJ7KEXCcC-OtdXKg,37
+sql_blocks/sql_blocks.py,sha256=qDwpII5rYY8tqpp93bKKHW9udOqI3SRb-0wfxjAsYD4,74694
+sql_blocks-1.25.6139999999999.dist-info/LICENSE,sha256=6kbiFSfobTZ7beWiKnHpN902HgBx-Jzgcme0SvKqhKY,1091
+sql_blocks-1.25.6139999999999.dist-info/METADATA,sha256=Pmk2nqJ3hV7PZVGE8V0bqw8TnSa1_rVIiylJad-nDCk,22236
+sql_blocks-1.25.6139999999999.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
+sql_blocks-1.25.6139999999999.dist-info/top_level.txt,sha256=57AbUvUjYNy4m1EqDaU3WHeP-uyIAfV0n8GAUp1a1YQ,11
+sql_blocks-1.25.6139999999999.dist-info/RECORD,,

sql_blocks-1.25.610999999999.dist-info/RECORD

@@ -1,7 +0,0 @@
-sql_blocks/__init__.py,sha256=5ItzGCyqqa6kwY8wvF9kapyHsAiWJ7KEXCcC-OtdXKg,37
-sql_blocks/sql_blocks.py,sha256=tdfGConHw2iosex_BSXAWxYTGufPyOVMSTyi3g6gqpM,71400
-sql_blocks-1.25.610999999999.dist-info/LICENSE,sha256=6kbiFSfobTZ7beWiKnHpN902HgBx-Jzgcme0SvKqhKY,1091
-sql_blocks-1.25.610999999999.dist-info/METADATA,sha256=09DIeKq_SinVNKt0OyYQQDOZ2_DDR6CmoP_n99t4ZXA,22235
-sql_blocks-1.25.610999999999.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
-sql_blocks-1.25.610999999999.dist-info/top_level.txt,sha256=57AbUvUjYNy4m1EqDaU3WHeP-uyIAfV0n8GAUp1a1YQ,11
-sql_blocks-1.25.610999999999.dist-info/RECORD,,