Improve Non_Unique_Primary_Key error, split file format detection into read/write, improve SQLite format detection

distribution/lib/Standard/Base/0.0.0-dev/src/Data/Array.enso

@@ -438,6 +438,8 @@ type Array
     flatten : Vector Any
     flatten self = Vector.flatten self
 
+    ## PRIVATE
+       ADVANCED
     short_display_text : Integer -> Text
     short_display_text self max_entries=10 = Vector.short_display_text self max_entries
 
@@ -641,9 +643,15 @@ type Array
     join : Text -> Text -> Text -> Text
     join self separator="" prefix="" suffix="" = Vector.join self separator prefix suffix
 
+    ## PRIVATE
+       Generates a human-readable text representation of the array.
     to_text : Text
     to_text self = self.map .to_text . join ", " "[" "]"
 
+    ## PRIVATE
+    to_display_text : Text
+    to_display_text self = self.short_display_text max_entries=40
+
     ## Combines all the elements of a non-empty array using a binary operation.
        If the array is empty, it returns `if_empty`.
 

distribution/lib/Standard/Base/0.0.0-dev/src/Data/Vector.enso

@@ -618,6 +618,10 @@ type Vector a
     to_text : Text
     to_text self = self.map .to_text . join ", " "[" "]"
 
+    ## PRIVATE
+    to_display_text : Text
+    to_display_text self = self.short_display_text max_entries=40
+
     ## PRIVATE
        ADVANCED
 

distribution/lib/Standard/Base/0.0.0-dev/src/System/File.enso

@@ -245,7 +245,8 @@ type File
     @format format_widget
     read : File_Format -> Problem_Behavior -> Any ! File_Error
     read self format=Auto_Detect (on_problems=Problem_Behavior.Report_Warning) =
-        format.read self on_problems
+        if self.exists.not then Error.throw (File_Error.Not_Found self) else
+            format.read self on_problems
 
     ## ALIAS Load Bytes, Open Bytes
        Reads all bytes in this file into a byte vector.
@@ -612,6 +613,14 @@ type File
             resource = Managed_Resource.register stream close_stream
             Input_Stream.Value self resource
 
+    ## PRIVATE
+       Reads first `n` bytes from the file (or less if the file is too small)
+       and returns a vector of bytes.
+    read_first_bytes : Integer -> Vector ! File_Error
+    read_first_bytes self n =
+        opts = [File_Access.Read]
+        self.with_input_stream opts (_.read_n_bytes n)
+
     ## PRIVATE
        Reads last `n` bytes from the file (or less if the file is too small) and
        returns a vector of bytes.

distribution/lib/Standard/Base/0.0.0-dev/src/System/File_Format.enso

@@ -58,14 +58,26 @@ type Auto_Detect
        Implements the `File.read` for this `File_Format`
     read : File -> Problem_Behavior -> Any ! File_Error
     read self file on_problems =
-        reader = Auto_Detect.get_format file
+        reader = Auto_Detect.get_reading_format file
         if reader == Nothing then Error.throw (File_Error.Unsupported_Type file) else
             reader.read file on_problems
 
     ## PRIVATE
-    get_format : File -> Any | Nothing
-    get_format file =
-        get_format f-> f.for_file file
+       Finds a matching format for reading the file.
+
+       It assumes that `file` already exists.
+    get_reading_format : File -> Any | Nothing
+    get_reading_format file =
+        get_format f-> f.for_file_read file
+
+    ## PRIVATE
+       Finds a matching format for reading the file.
+
+       It may not assume that the `file` exists, so it must only rely on the
+       file path (extension in particular), but not the contents.
+    get_writing_format : File -> Any | Nothing
+    get_writing_format file =
+        get_format f-> f.for_file_write file
 
     ## PRIVATE
     get_web_parser : Text -> URI -> Any | Nothing
@@ -91,13 +103,18 @@ type Plain_Text_Format
 
     ## PRIVATE
        If the File_Format supports reading from the file, return a configured instance.
-    for_file : File -> Plain_Text_Format | Nothing
-    for_file file =
+    for_file_read : File -> Plain_Text_Format | Nothing
+    for_file_read file =
         case file.extension of
             ".txt" -> Plain_Text_Format.Plain_Text
             ".log" -> Plain_Text_Format.Plain_Text
             _ -> Nothing
 
+    ## PRIVATE
+       If this File_Format should be used for writing to that file, return a configured instance.
+    for_file_write : File -> Plain_Text_Format | Nothing
+    for_file_write file = Plain_Text_Format.for_file_read file
+
     ## PRIVATE
        If the File_Format supports reading from the web response, return a configured instance.
     for_web : Text -> URI -> Plain_Text_Format | Nothing
@@ -127,12 +144,17 @@ type Plain_Text_Format
 type Bytes
     ## PRIVATE
        If the File_Format supports reading from the file, return a configured instance.
-    for_file : File -> Bytes | Nothing
-    for_file file =
+    for_file_read : File -> Bytes | Nothing
+    for_file_read file =
         case file.extension of
             ".dat" -> Bytes
             _ -> Nothing
 
+    ## PRIVATE
+       If this File_Format should be used for writing to that file, return a configured instance.
+    for_file_write : File -> Bytes | Nothing
+    for_file_write file = Bytes.for_file_read file
+
     ## PRIVATE
        If the File_Format supports reading from the web response, return a configured instance.
        As `Bytes`, does not support reading from the web returns `Nothing`.
@@ -148,13 +170,18 @@ type Bytes
 type JSON_Format
     ## PRIVATE
        If the File_Format supports reading from the file, return a configured instance.
-    for_file : File -> JSON_Format | Nothing
-    for_file file =
+    for_file_read : File -> JSON_Format | Nothing
+    for_file_read file =
         case file.extension of
             ".json" -> JSON_Format
             ".geojson" -> JSON_Format
             _ -> Nothing
 
+    ## PRIVATE
+       If this File_Format should be used for writing to that file, return a configured instance.
+    for_file_write : File -> JSON_Format | Nothing
+    for_file_write file = JSON_Format.for_file_read file
+
     ## PRIVATE
        If the File_Format supports reading from the web response, return a configured instance.
     for_web : Text -> URI -> JSON_Format | Nothing

distribution/lib/Standard/Database/0.0.0-dev/src/Connection/SQLite_Format.enso

@@ -11,8 +11,16 @@ type SQLite_Format
 
     ## PRIVATE
        If the File_Format supports reading from the file, return a configured instance.
-    for_file : File -> SQLite_Format | Nothing
-    for_file file =
+    for_file_read : File -> SQLite_Format | Nothing
+    for_file_read file =
+        expected_header = magic_header_string
+        got_header = file.read_first_bytes expected_header.length
+        if got_header == expected_header then SQLite_Format.For_File else Nothing
+
+    ## PRIVATE
+       If the File_Format supports writing to the file, return a configured instance.
+    for_file_write : File -> SQLite_Format | Nothing
+    for_file_write file =
         case file.extension of
             ".db" -> SQLite_Format.For_File
             ".sqlite" -> SQLite_Format.For_File
@@ -31,3 +39,8 @@ type SQLite_Format
     read self file on_problems =
         _ = [on_problems]
         Database.connect (SQLite_Details.SQLite file)
+
+## PRIVATE
+   Based on the File Format definition at: https://www.sqlite.org/fileformat.html
+magic_header_string =
+    "SQLite format 3".utf_8 + [0]

distribution/lib/Standard/Database/0.0.0-dev/src/Errors.enso

@@ -155,10 +155,14 @@ type Non_Unique_Primary_Key
 
        Arguments:
        - primary_key: The primary key that is not unique.
-    Error (primary_key : Vector Text)
+       - clashing_primary_key: The values of an example key that corresponds to
+         more than one row.
+       - clashing_example_row_count: The number of rows that correspond to the
+         example key.
+    Error (primary_key : Vector Text) (clashing_primary_key : Vector Any) (clashing_example_row_count : Integer)
 
     ## PRIVATE
        Pretty print the non-unique primary key error.
     to_display_text : Text
     to_display_text self =
-        "The primary key " + self.primary_key.to_display_text + " is not unique."
+        "The primary key " + self.primary_key.to_display_text + " is not unique. The key "+self.clashing_primary_key.to_display_text+" corresponds to "+self.clashing_example_row_count.to_text+" rows."

distribution/lib/Standard/Database/0.0.0-dev/src/Extensions/Upload_Table.enso

@@ -62,7 +62,7 @@ In_Memory_Table.create_database_table self connection table_name=Nothing primary
        continue. Otherwise, they could 'leak' to `Panic.rethrow` and be wrongly
        raised as panics.
     upload_status = create_table_statement.if_not_error <|
-        translate_known_upload_errors connection resolved_primary_key <|
+        translate_known_upload_errors self connection resolved_primary_key <|
             connection.jdbc_connection.run_within_transaction <|
                 Panic.rethrow <| connection.execute_update create_table_statement
                 if structure_only.not then
@@ -119,7 +119,7 @@ Database_Table.create_database_table self connection table_name=Nothing primary_
         Error.throw (Unsupported_Database_Operation.Error "The Database table to be uploaded must be coming from the same connection as the connection on which the new table is being created. Cross-connection uploads are currently not supported. To work around this, you can first `.read` the table into memory and then upload it from memory to a different connection.")
 
     upload_status = connection_check.if_not_error <| create_table_statement.if_not_error <|
-        translate_known_upload_errors connection resolved_primary_key <|
+        translate_known_upload_errors self connection resolved_primary_key <|
             connection.jdbc_connection.run_within_transaction <|
                 Panic.rethrow <| connection.execute_update create_table_statement
                 if structure_only.not then
@@ -144,15 +144,35 @@ resolve_primary_key table primary_key = case primary_key of
 ## PRIVATE
    Inspects any `SQL_Error` thrown and replaces it with a more precise error
    type when available.
-translate_known_upload_errors connection primary_key ~action =
+translate_known_upload_errors source_table connection primary_key ~action =
     handler caught_panic =
         error_mapper = connection.dialect.get_error_mapper
         sql_error = caught_panic.payload
         case error_mapper.is_primary_key_violation sql_error of
-            True -> Error.throw (Non_Unique_Primary_Key.Error primary_key)
+            True -> raise_duplicated_primary_key_error source_table primary_key caught_panic
             False -> Panic.throw caught_panic
     Panic.catch SQL_Error action handler
 
+## PRIVATE
+   Creates a `Non_Unique_Primary_Key` error containing information about an
+   example group violating the uniqueness constraint.
+raise_duplicated_primary_key_error source_table primary_key original_panic =
+    agg = source_table.aggregate [Aggregate_Column.Count]+(primary_key.map Aggregate_Column.Group_By)
+    filtered = agg.filter column=0 (Filter_Condition.Greater than=1)
+    materialized = filtered.read max_rows=1
+    case materialized.row_count == 0 of
+        ## If we couldn't find a duplicated key, we give up the translation and
+           rethrow the original panic containing the SQL error. This could
+           happen if the constraint violation is on some non-trivial key, like
+           case insensitive.
+        True -> Panic.throw original_panic
+        False ->
+            row = materialized.first_row.to_vector
+            example_count = row.first
+            example_entry = row.drop 1
+            Error.throw (Non_Unique_Primary_Key.Error primary_key example_entry example_count)
+
+
 ## PRIVATE
    Creates a statement that will create a table with structure determined by the
    provided columns.

distribution/lib/Standard/Image/0.0.0-dev/src/Image_File_Format.enso

@@ -14,11 +14,16 @@ type Image_File_Format
 
     ## PRIVATE
        If the File_Format supports reading from the file, return a configured instance.
-    for_file : File -> Image_File_Format | Nothing
-    for_file file =
+    for_file_read : File -> Image_File_Format | Nothing
+    for_file_read file =
         extension = file.extension
         if supported.contains extension then Image_File_Format.For_File else Nothing
 
+    ## PRIVATE
+       If this File_Format should be used for writing to that file, return a configured instance.
+    for_file_write : File -> Image_File_Format | Nothing
+    for_file_write file = Image_File_Format.for_file_read file
+
     ## PRIVATE
        If the File_Format supports reading from the web response, return a configured instance.
     for_web : Text -> URI -> Image_File_Format | Nothing

distribution/lib/Standard/Table/0.0.0-dev/src/Data/Table.enso

@@ -1660,6 +1660,21 @@ type Table
     row_count : Integer
     row_count self = self.java_table.rowCount
 
+    ## Returns a materialized dataframe containing rows of this table.
+
+       In the in-memory backend, this returns the same table, truncated to
+       `max_rows`. This is only kept for API compatibility between database and
+       in-memory tables. The `read` operation can be used to ensure that the
+       table is now in-memory, regardless of its origin.
+
+       Arguments:
+       - max_rows: specifies a maximum amount of rows to fetch; if not set, all
+         available rows are fetched.
+    read : (Integer | Nothing) -> Table
+    read self max_rows=Nothing = case max_rows of
+        Nothing -> self
+        _ : Integer -> self.take (First max_rows)
+
     ## Returns a Table describing this table's contents.
 
        The table lists all columns, counts of non-null items and value types of
@@ -1913,7 +1928,7 @@ type Table
         file = File.new path
         case format of
             _ : Auto_Detect ->
-                base_format = format.get_format file
+                base_format = format.get_writing_format file
                 if base_format == Nothing then Error.throw (File_Error.Unsupported_Output_Type file Table) else
                     self.write file format=base_format on_existing_file match_columns on_problems
             _ ->

distribution/lib/Standard/Table/0.0.0-dev/src/Delimited/Delimited_Format.enso

@@ -54,14 +54,19 @@ type Delimited_Format
     ## PRIVATE
        ADVANCED
        If the File_Format supports reading from the file, return a configured instance.
-    for_file : File -> Delimited_Format | Nothing
-    for_file file =
+    for_file_read : File -> Delimited_Format | Nothing
+    for_file_read file =
         case file.extension of
             ".csv" -> Delimited_Format.Delimited ','
             ".tab" -> Delimited_Format.Delimited '\t'
             ".tsv" -> Delimited_Format.Delimited '\t'
             _ -> Nothing
 
+    ## PRIVATE
+       If this File_Format should be used for writing to that file, return a configured instance.
+    for_file_write : File -> Delimited_Format | Nothing
+    for_file_write file = Delimited_Format.for_file_read file
+
     ## PRIVATE
        ADVANCED
        If the File_Format supports reading from the web response, return a configured instance.

distribution/lib/Standard/Table/0.0.0-dev/src/Errors.enso

@@ -12,6 +12,13 @@ polyglot java import org.enso.table.error.EmptySheetException
 type Missing_Input_Columns
     ## PRIVATE
        One or more columns not found in the input table.
+
+       Arguments:
+       - criteria: the names of the columns or regular expressions that did not
+         have any matches.
+       - where: an optional text describing to which object this error is
+         related to (for example in join, whether the reported error is for the
+         left or right table).
     Error (criteria : [Text]) (where:Text|Nothing = Nothing)
 
     ## PRIVATE

distribution/lib/Standard/Table/0.0.0-dev/src/Excel/Excel_Format.enso

@@ -48,12 +48,17 @@ type Excel_Format
     ## PRIVATE
        ADVANCED
        If the File_Format supports reading from the file, return a configured instance.
-    for_file : File -> Excel_Format | Nothing
-    for_file file =
+    for_file_read : File -> Excel_Format | Nothing
+    for_file_read file =
         is_xls = should_treat_as_xls_format Infer file
         if is_xls.is_error then Nothing else
             Excel_Format.Excel xls_format=is_xls
 
+    ## PRIVATE
+       If this File_Format should be used for writing to that file, return a configured instance.
+    for_file_write : File -> Excel_Format | Nothing
+    for_file_write file = Excel_Format.for_file_read file
+
     ## PRIVATE
        ADVANCED
        If the File_Format supports reading from the web response, return a configured instance.

distribution/lib/Standard/Test/0.0.0-dev/src/Extensions.enso

@@ -284,7 +284,8 @@ Error.should_succeed self frames_to_skip=0 =
 
 ## Handles an unexpected dataflow error.
 Error.should_be_a : Integer -> Any
-Error.should_be_a self frames_to_skip=0 =
+Error.should_be_a self typ frames_to_skip=0 =
+    _ = typ
     Test.fail_match_on_unexpected_error self 1+frames_to_skip
 
 ## Asserts that the given `Boolean` is `True`

test/Image_Tests/src/Image_Read_Write_Spec.enso

@@ -69,9 +69,9 @@ spec =
 
     Test.group "Image File_Format" <|
         Test.specify "should recognise image files" <|
-            Auto_Detect.get_format (enso_project.data / "data.jpg") . should_be_a Image_File_Format
-            Auto_Detect.get_format (enso_project.data / "data.png") . should_be_a Image_File_Format
-            Auto_Detect.get_format (enso_project.data / "data.bmp") . should_be_a Image_File_Format
+            Auto_Detect.get_reading_format (enso_project.data / "data.jpg") . should_be_a Image_File_Format
+            Auto_Detect.get_reading_format (enso_project.data / "data.png") . should_be_a Image_File_Format
+            Auto_Detect.get_reading_format (enso_project.data / "data.bmp") . should_be_a Image_File_Format
 
         Test.specify "should allow reading an Image" <|
             img = Data.read rgba_file