polars-df 0.6.0 → 0.7.0

data/lib/polars/string_expr.rb

@@ -82,9 +82,13 @@ module Polars
       time_zone: nil,
       strict: true,
       exact: true,
-      cache: true
+      cache: true,
+      use_earliest: nil,
+      ambiguous: "raise"
     )
       _validate_format_argument(format)
+      ambiguous = Utils.rename_use_earliest_to_ambiguous(use_earliest, ambiguous)
+      ambiguous = Polars.lit(ambiguous) unless ambiguous.is_a?(Expr)
       Utils.wrap_expr(
         self._rbexpr.str_to_datetime(
           format,
@@ -92,7 +96,8 @@ module Polars
           time_zone,
           strict,
           exact,
-          cache
+          cache,
+          ambiguous._rbexpr
         )
       )
     end
@@ -235,7 +240,7 @@ module Polars
     #   # │ 東京 ┆ 6      ┆ 2      │
     #   # └──────┴────────┴────────┘
     def lengths
-      Utils.wrap_expr(_rbexpr.str_lengths)
+      Utils.wrap_expr(_rbexpr.str_len_bytes)
     end
 
     # Get length of the strings as `:u32` (as number of chars).
@@ -267,13 +272,15 @@ module Polars
     #   # │ 東京 ┆ 6      ┆ 2      │
     #   # └──────┴────────┴────────┘
     def n_chars
-      Utils.wrap_expr(_rbexpr.str_n_chars)
+      Utils.wrap_expr(_rbexpr.str_len_chars)
     end
 
     # Vertically concat the values in the Series to a single string value.
     #
     # @param delimiter [String]
     #   The delimiter to insert between consecutive string values.
+    # @param ignore_nulls [Boolean]
+    #   Ignore null values (default).
     #
     # @return [Expr]
     #
@@ -282,15 +289,28 @@ module Polars
     #   df.select(Polars.col("foo").str.concat("-"))
     #   # =>
     #   # shape: (1, 1)
-    #   # ┌──────────┐
-    #   # │ foo      │
-    #   # │ ---      │
-    #   # │ str      │
-    #   # ╞══════════╡
-    #   # │ 1-null-2 │
-    #   # └──────────┘
-    def concat(delimiter = "-")
-      Utils.wrap_expr(_rbexpr.str_concat(delimiter))
+    #   # ┌─────┐
+    #   # │ foo │
+    #   # │ --- │
+    #   # │ str │
+    #   # ╞═════╡
+    #   # │ 1-2 │
+    #   # └─────┘
+    #
+    # @example
+    #   df = Polars::DataFrame.new({"foo" => [1, nil, 2]})
+    #   df.select(Polars.col("foo").str.concat("-", ignore_nulls: false))
+    #   # =>
+    #   # shape: (1, 1)
+    #   # ┌──────┐
+    #   # │ foo  │
+    #   # │ ---  │
+    #   # │ str  │
+    #   # ╞══════╡
+    #   # │ null │
+    #   # └──────┘
+    def concat(delimiter = "-", ignore_nulls: true)
+      Utils.wrap_expr(_rbexpr.str_concat(delimiter, ignore_nulls))
     end
 
     # Transform to uppercase variant.
@@ -337,7 +357,7 @@ module Polars
 
     # Remove leading and trailing whitespace.
     #
-    # @param matches [String, nil]
+    # @param characters [String, nil]
     #   An optional single character that should be trimmed.
     #
     # @return [Expr]
@@ -356,16 +376,15 @@ module Polars
     #   # │ trail │
     #   # │ both  │
     #   # └───────┘
-    def strip(matches = nil)
-      if !matches.nil? && matches.length > 1
-        raise ArgumentError, "matches should contain a single character"
-      end
-      Utils.wrap_expr(_rbexpr.str_strip(matches))
+    def strip_chars(characters = nil)
+      characters = Utils.parse_as_expression(characters, str_as_lit: true)
+      Utils.wrap_expr(_rbexpr.str_strip_chars(characters))
     end
+    alias_method :strip, :strip_chars
 
     # Remove leading whitespace.
     #
-    # @param matches [String, nil]
+    # @param characters [String, nil]
     #   An optional single character that should be trimmed.
     #
     # @return [Expr]
@@ -384,16 +403,15 @@ module Polars
     #   # │ trail  │
     #   # │ both   │
     #   # └────────┘
-    def lstrip(matches = nil)
-      if !matches.nil? && matches.length > 1
-        raise ArgumentError, "matches should contain a single character"
-      end
-      Utils.wrap_expr(_rbexpr.str_lstrip(matches))
+    def strip_chars_start(characters = nil)
+      characters = Utils.parse_as_expression(characters, str_as_lit: true)
+      Utils.wrap_expr(_rbexpr.str_strip_chars_start(characters))
     end
+    alias_method :lstrip, :strip_chars_start
 
     # Remove trailing whitespace.
     #
-    # @param matches [String, nil]
+    # @param characters [String, nil]
     #   An optional single character that should be trimmed.
     #
     # @return [Expr]
@@ -412,12 +430,11 @@ module Polars
     #   # │ trail │
     #   # │  both │
     #   # └───────┘
-    def rstrip(matches = nil)
-      if !matches.nil? && matches.length > 1
-        raise ArgumentError, "matches should contain a single character"
-      end
-      Utils.wrap_expr(_rbexpr.str_rstrip(matches))
+    def strip_chars_end(characters = nil)
+      characters = Utils.parse_as_expression(characters, str_as_lit: true)
+      Utils.wrap_expr(_rbexpr.str_strip_chars_end(characters))
     end
+    alias_method :rstrip, :strip_chars_end
 
     # Fills the string with zeroes.
     #
@@ -461,13 +478,13 @@ module Polars
       Utils.wrap_expr(_rbexpr.str_zfill(alignment))
     end
 
-    # Return the string left justified in a string of length `width`.
+    # Return the string left justified in a string of length `length`.
     #
     # Padding is done using the specified `fillchar`.
-    # The original string is returned if `width` is less than or equal to
+    # The original string is returned if `length` is less than or equal to
     # `s.length`.
     #
-    # @param width [Integer]
+    # @param length [Integer]
     #   Justify left to this length.
     # @param fillchar [String]
     #   Fill with this ASCII character.
@@ -489,17 +506,18 @@ module Polars
     #   # │ null         │
     #   # │ hippopotamus │
     #   # └──────────────┘
-    def ljust(width, fillchar = " ")
-      Utils.wrap_expr(_rbexpr.str_ljust(width, fillchar))
+    def ljust(length, fillchar = " ")
+      Utils.wrap_expr(_rbexpr.str_pad_end(length, fillchar))
     end
+    alias_method :pad_end, :ljust
 
-    # Return the string right justified in a string of length `width`.
+    # Return the string right justified in a string of length `length`.
     #
     # Padding is done using the specified `fillchar`.
-    # The original string is returned if `width` is less than or equal to
+    # The original string is returned if `length` is less than or equal to
     # `s.length`.
     #
-    # @param width [Integer]
+    # @param length [Integer]
     #   Justify right to this length.
     # @param fillchar [String]
     #   Fill with this ASCII character.
@@ -521,9 +539,10 @@ module Polars
     #   # │ null         │
     #   # │ hippopotamus │
     #   # └──────────────┘
-    def rjust(width, fillchar = " ")
-      Utils.wrap_expr(_rbexpr.str_rjust(width, fillchar))
+    def rjust(length, fillchar = " ")
+      Utils.wrap_expr(_rbexpr.str_pad_start(length, fillchar))
     end
+    alias_method :pad_start, :rjust
 
     # Check if string contains a substring that matches a regex.
     #
@@ -864,9 +883,11 @@ module Polars
     #   # │ 5            │
     #   # │ 6            │
     #   # └──────────────┘
-    def count_match(pattern)
-      Utils.wrap_expr(_rbexpr.count_match(pattern))
+    def count_matches(pattern, literal: false)
+      pattern = Utils.parse_as_expression(pattern, str_as_lit: true)
+      Utils.wrap_expr(_rbexpr.str_count_matches(pattern, literal))
     end
+    alias_method :count_match, :count_matches
 
     # Split the string by a substring.
     #
@@ -892,6 +913,7 @@ module Polars
     #   # │ ["foo", "bar", "baz"] │
     #   # └───────────────────────┘
     def split(by, inclusive: false)
+      by = Utils.parse_as_expression(by, str_as_lit: true)
       if inclusive
         Utils.wrap_expr(_rbexpr.str_split_inclusive(by))
       else
@@ -934,6 +956,7 @@ module Polars
     #   # │ {"d","4"}   │
     #   # └─────────────┘
     def split_exact(by, n, inclusive: false)
+      by = Utils.parse_as_expression(by, str_as_lit: true)
       if inclusive
         Utils.wrap_expr(_rbexpr.str_split_exact_inclusive(by, n))
       else
@@ -970,6 +993,7 @@ module Polars
     #   # │ {"foo","bar baz"} │
     #   # └───────────────────┘
     def splitn(by, n)
+      by = Utils.parse_as_expression(by, str_as_lit: true)
       Utils.wrap_expr(_rbexpr.str_splitn(by, n))
     end
 
@@ -1091,6 +1115,52 @@ module Polars
       Utils.wrap_expr(_rbexpr.str_explode)
     end
 
+    # Convert an Utf8 column into an Int64 column with base radix.
+    #
+    # @param base [Integer]
+    #   Positive integer which is the base of the string we are parsing.
+    #   Default: 10.
+    # @param strict [Boolean]
+    #   Bool, default=true will raise any ParseError or overflow as ComputeError.
+    #   false silently convert to Null.
+    #
+    # @return [Expr]
+    #
+    # @example
+    #   df = Polars::DataFrame.new({"bin" => ["110", "101", "010", "invalid"]})
+    #   df.with_columns(Polars.col("bin").str.to_integer(base: 2, strict: false).alias("parsed"))
+    #   # =>
+    #   # shape: (4, 2)
+    #   # ┌─────────┬────────┐
+    #   # │ bin     ┆ parsed │
+    #   # │ ---     ┆ ---    │
+    #   # │ str     ┆ i64    │
+    #   # ╞═════════╪════════╡
+    #   # │ 110     ┆ 6      │
+    #   # │ 101     ┆ 5      │
+    #   # │ 010     ┆ 2      │
+    #   # │ invalid ┆ null   │
+    #   # └─────────┴────────┘
+    #
+    # @example
+    #   df = Polars::DataFrame.new({"hex" => ["fa1e", "ff00", "cafe", nil]})
+    #   df.with_columns(Polars.col("hex").str.to_integer(base: 16, strict: true).alias("parsed"))
+    #   # =>
+    #   # shape: (4, 2)
+    #   # ┌──────┬────────┐
+    #   # │ hex  ┆ parsed │
+    #   # │ ---  ┆ ---    │
+    #   # │ str  ┆ i64    │
+    #   # ╞══════╪════════╡
+    #   # │ fa1e ┆ 64030  │
+    #   # │ ff00 ┆ 65280  │
+    #   # │ cafe ┆ 51966  │
+    #   # │ null ┆ null   │
+    #   # └──────┴────────┘
+    def to_integer(base: 10, strict: true)
+      Utils.wrap_expr(_rbexpr.str_to_integer(base, strict))
+    end
+
     # Parse integers with base radix from strings.
     #
     # By default base 2. ParseError/Overflows become Nulls.
@@ -1119,24 +1189,8 @@ module Polars
     #   # │ 2    │
     #   # │ null │
     #   # └──────┘
-    #
-    # @example
-    #   df = Polars::DataFrame.new({"hex" => ["fa1e", "ff00", "cafe", nil]})
-    #   df.select(Polars.col("hex").str.parse_int(16, strict: true))
-    #   # =>
-    #   # shape: (4, 1)
-    #   # ┌───────┐
-    #   # │ hex   │
-    #   # │ ---   │
-    #   # │ i32   │
-    #   # ╞═══════╡
-    #   # │ 64030 │
-    #   # │ 65280 │
-    #   # │ 51966 │
-    #   # │ null  │
-    #   # └───────┘
     def parse_int(radix = 2, strict: true)
-      Utils.wrap_expr(_rbexpr.str_parse_int(radix, strict))
+      to_integer(base: 2, strict: strict).cast(Int32, strict: strict)
     end
 
     private

data/lib/polars/string_name_space.rb

@@ -82,7 +82,8 @@ module Polars
       time_zone: nil,
       strict: true,
       exact: true,
-      cache: true
+      cache: true,
+      use_earliest: nil
     )
       super
     end
@@ -232,9 +233,23 @@ module Polars
     # @return [Series]
     #
     # @example
-    #   Polars::Series.new([1, nil, 2]).str.concat("-")[0]
-    #   # => "1-null-2"
-    def concat(delimiter = "-")
+    #   Polars::Series.new([1, nil, 2]).str.concat("-")
+    #   # =>
+    #   # shape: (1,)
+    #   # Series: '' [str]
+    #   # [
+    #   #         "1-2"
+    #   # ]
+    #
+    # @example
+    #   Polars::Series.new([1, nil, 2]).str.concat("-", ignore_nulls: false)
+    #   # =>
+    #   # shape: (1,)
+    #   # Series: '' [str]
+    #   # [
+    #   #         null
+    #   # ]
+    def concat(delimiter = "-", ignore_nulls: true)
       super
     end
 

data/lib/polars/utils.rb

@@ -308,5 +308,17 @@ module Polars
 
       expr._rbexpr
     end
+
+    USE_EARLIEST_TO_AMBIGUOUS = {
+      true => "earliest",
+      false => "latest"
+    }
+
+    def self.rename_use_earliest_to_ambiguous(use_earliest, ambiguous)
+      unless use_earliest.nil?
+        ambiguous = USE_EARLIEST_TO_AMBIGUOUS.fetch(use_earliest)
+      end
+      ambiguous
+    end
   end
 end

data/lib/polars/version.rb

@@ -1,4 +1,4 @@
 module Polars
   # @private
-  VERSION = "0.6.0"
+  VERSION = "0.7.0"
 end

data/lib/polars.rb

@@ -19,6 +19,7 @@ require_relative "polars/binary_expr"
 require_relative "polars/binary_name_space"
 require_relative "polars/cat_expr"
 require_relative "polars/cat_name_space"
+require_relative "polars/config"
 require_relative "polars/convert"
 require_relative "polars/plot"
 require_relative "polars/data_frame"
@@ -37,9 +38,11 @@ require_relative "polars/lazy_group_by"
 require_relative "polars/list_expr"
 require_relative "polars/list_name_space"
 require_relative "polars/meta_expr"
+require_relative "polars/name_expr"
 require_relative "polars/rolling_group_by"
 require_relative "polars/series"
 require_relative "polars/slice"
+require_relative "polars/sql_context"
 require_relative "polars/string_expr"
 require_relative "polars/string_name_space"
 require_relative "polars/struct_expr"

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: polars-df
 version: !ruby/object:Gem::Version
-  version: 0.6.0
+  version: 0.7.0
 platform: ruby
 authors:
 - Andrew Kane
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2023-07-24 00:00:00.000000000 Z
+date: 2023-11-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rb_sys
@@ -39,15 +39,10 @@ files:
 - README.md
 - ext/polars/Cargo.toml
 - ext/polars/extconf.rb
-- ext/polars/src/apply/dataframe.rs
-- ext/polars/src/apply/lazy.rs
-- ext/polars/src/apply/mod.rs
-- ext/polars/src/apply/series.rs
 - ext/polars/src/batched_csv.rs
 - ext/polars/src/conversion.rs
 - ext/polars/src/dataframe.rs
 - ext/polars/src/error.rs
-- ext/polars/src/expr.rs
 - ext/polars/src/expr/array.rs
 - ext/polars/src/expr/binary.rs
 - ext/polars/src/expr/categorical.rs
@@ -55,18 +50,30 @@ files:
 - ext/polars/src/expr/general.rs
 - ext/polars/src/expr/list.rs
 - ext/polars/src/expr/meta.rs
+- ext/polars/src/expr/mod.rs
+- ext/polars/src/expr/name.rs
+- ext/polars/src/expr/rolling.rs
 - ext/polars/src/expr/string.rs
 - ext/polars/src/expr/struct.rs
 - ext/polars/src/file.rs
+- ext/polars/src/functions/aggregation.rs
 - ext/polars/src/functions/eager.rs
 - ext/polars/src/functions/io.rs
 - ext/polars/src/functions/lazy.rs
 - ext/polars/src/functions/meta.rs
+- ext/polars/src/functions/misc.rs
 - ext/polars/src/functions/mod.rs
+- ext/polars/src/functions/random.rs
+- ext/polars/src/functions/range.rs
+- ext/polars/src/functions/string_cache.rs
 - ext/polars/src/functions/whenthen.rs
 - ext/polars/src/lazyframe.rs
 - ext/polars/src/lazygroupby.rs
 - ext/polars/src/lib.rs
+- ext/polars/src/map/dataframe.rs
+- ext/polars/src/map/lazy.rs
+- ext/polars/src/map/mod.rs
+- ext/polars/src/map/series.rs
 - ext/polars/src/object.rs
 - ext/polars/src/prelude.rs
 - ext/polars/src/rb_modules.rs
@@ -77,6 +84,7 @@ files:
 - ext/polars/src/series/construction.rs
 - ext/polars/src/series/export.rs
 - ext/polars/src/series/set_at_idx.rs
+- ext/polars/src/sql.rs
 - ext/polars/src/utils.rs
 - lib/polars-df.rb
 - lib/polars.rb
@@ -87,6 +95,7 @@ files:
 - lib/polars/binary_name_space.rb
 - lib/polars/cat_expr.rb
 - lib/polars/cat_name_space.rb
+- lib/polars/config.rb
 - lib/polars/convert.rb
 - lib/polars/data_frame.rb
 - lib/polars/data_types.rb
@@ -105,10 +114,12 @@ files:
 - lib/polars/list_expr.rb
 - lib/polars/list_name_space.rb
 - lib/polars/meta_expr.rb
+- lib/polars/name_expr.rb
 - lib/polars/plot.rb
 - lib/polars/rolling_group_by.rb
 - lib/polars/series.rb
 - lib/polars/slice.rb
+- lib/polars/sql_context.rb
 - lib/polars/string_expr.rb
 - lib/polars/string_name_space.rb
 - lib/polars/struct_expr.rb