prism 1.9.0 → 2.0.0.rc2

data/include/prism/util/pm_integer.h

@@ -1,130 +0,0 @@
-/**
- * @file pm_integer.h
- *
- * This module provides functions for working with arbitrary-sized integers.
- */
-#ifndef PRISM_NUMBER_H
-#define PRISM_NUMBER_H
-
-#include "prism/defines.h"
-#include "prism/util/pm_buffer.h"
-
-#include <assert.h>
-#include <stdbool.h>
-#include <stdint.h>
-#include <stdlib.h>
-
-/**
- * A structure represents an arbitrary-sized integer.
- */
-typedef struct {
-    /**
-     * The number of allocated values. length is set to 0 if the integer fits
-     * into uint32_t.
-     */
-    size_t length;
-
-    /**
-     * List of 32-bit integers. Set to NULL if the integer fits into uint32_t.
-     */
-    uint32_t *values;
-
-    /**
-     * Embedded value for small integer. This value is set to 0 if the value
-     * does not fit into uint32_t.
-     */
-    uint32_t value;
-
-    /**
-     * Whether or not the integer is negative. It is stored this way so that a
-     * zeroed pm_integer_t is always positive zero.
-     */
-    bool negative;
-} pm_integer_t;
-
-/**
- * An enum controlling the base of an integer. It is expected that the base is
- * already known before parsing the integer, even though it could be derived
- * from the string itself.
- */
-typedef enum {
-    /** The default decimal base, with no prefix. Leading 0s will be ignored. */
-    PM_INTEGER_BASE_DEFAULT,
-
-    /** The binary base, indicated by a 0b or 0B prefix. */
-    PM_INTEGER_BASE_BINARY,
-
-    /** The octal base, indicated by a 0, 0o, or 0O prefix. */
-    PM_INTEGER_BASE_OCTAL,
-
-    /** The decimal base, indicated by a 0d, 0D, or empty prefix. */
-    PM_INTEGER_BASE_DECIMAL,
-
-    /** The hexadecimal base, indicated by a 0x or 0X prefix. */
-    PM_INTEGER_BASE_HEXADECIMAL,
-
-    /**
-     * An unknown base, in which case pm_integer_parse will derive it based on
-     * the content of the string. This is less efficient and does more
-     * comparisons, so if callers know the base ahead of time, they should use
-     * that instead.
-     */
-    PM_INTEGER_BASE_UNKNOWN
-} pm_integer_base_t;
-
-/**
- * Parse an integer from a string. This assumes that the format of the integer
- * has already been validated, as internal validation checks are not performed
- * here.
- *
- * @param integer The integer to parse into.
- * @param base The base of the integer.
- * @param start The start of the string.
- * @param end The end of the string.
- */
-void pm_integer_parse(pm_integer_t *integer, pm_integer_base_t base, const uint8_t *start, const uint8_t *end);
-
-/**
- * Compare two integers. This function returns -1 if the left integer is less
- * than the right integer, 0 if they are equal, and 1 if the left integer is
- * greater than the right integer.
- *
- * @param left The left integer to compare.
- * @param right The right integer to compare.
- * @return The result of the comparison.
- */
-int pm_integer_compare(const pm_integer_t *left, const pm_integer_t *right);
-
-/**
- * Reduce a ratio of integers to its simplest form.
- *
- * If either the numerator or denominator do not fit into a 32-bit integer, then
- * this function is a no-op. In the future, we may consider reducing even the
- * larger numbers, but for now we're going to keep it simple.
- *
- * @param numerator The numerator of the ratio.
- * @param denominator The denominator of the ratio.
- */
-void pm_integers_reduce(pm_integer_t *numerator, pm_integer_t *denominator);
-
-/**
- * Convert an integer to a decimal string.
- *
- * @param buffer The buffer to append the string to.
- * @param integer The integer to convert to a string.
- *
- * \public \memberof pm_integer_t
- */
-PRISM_EXPORTED_FUNCTION void pm_integer_string(pm_buffer_t *buffer, const pm_integer_t *integer);
-
-/**
- * Free the internal memory of an integer. This memory will only be allocated if
- * the integer exceeds the size of a single node in the linked list.
- *
- * @param integer The integer to free.
- *
- * \public \memberof pm_integer_t
- */
-PRISM_EXPORTED_FUNCTION void pm_integer_free(pm_integer_t *integer);
-
-#endif

data/include/prism/util/pm_memchr.h

@@ -1,29 +0,0 @@
-/**
- * @file pm_memchr.h
- *
- * A custom memchr implementation.
- */
-#ifndef PRISM_MEMCHR_H
-#define PRISM_MEMCHR_H
-
-#include "prism/defines.h"
-#include "prism/encoding.h"
-
-#include <stddef.h>
-
-/**
- * We need to roll our own memchr to handle cases where the encoding changes and
- * we need to search for a character in a buffer that could be the trailing byte
- * of a multibyte character.
- *
- * @param source The source string.
- * @param character The character to search for.
- * @param number The maximum number of bytes to search.
- * @param encoding_changed Whether the encoding changed.
- * @param encoding A pointer to the encoding.
- * @return A pointer to the first occurrence of the character in the source
- *     string, or NULL if no such character exists.
- */
-void * pm_memchr(const void *source, int character, size_t number, bool encoding_changed, const pm_encoding_t *encoding);
-
-#endif

data/include/prism/util/pm_newline_list.h

@@ -1,113 +0,0 @@
-/**
- * @file pm_newline_list.h
- *
- * A list of byte offsets of newlines in a string.
- *
- * When compiling the syntax tree, it's necessary to know the line and column
- * of many nodes. This is necessary to support things like error messages,
- * tracepoints, etc.
- *
- * It's possible that we could store the start line, start column, end line, and
- * end column on every node in addition to the offsets that we already store,
- * but that would be quite a lot of memory overhead.
- */
-#ifndef PRISM_NEWLINE_LIST_H
-#define PRISM_NEWLINE_LIST_H
-
-#include "prism/defines.h"
-
-#include <assert.h>
-#include <stdbool.h>
-#include <stddef.h>
-#include <stdlib.h>
-
-/**
- * A list of offsets of newlines in a string. The offsets are assumed to be
- * sorted/inserted in ascending order.
- */
-typedef struct {
-    /** A pointer to the start of the source string. */
-    const uint8_t *start;
-
-    /** The number of offsets in the list. */
-    size_t size;
-
-    /** The capacity of the list that has been allocated. */
-    size_t capacity;
-
-    /** The list of offsets. */
-    size_t *offsets;
-} pm_newline_list_t;
-
-/**
- * A line and column in a string.
- */
-typedef struct {
-    /** The line number. */
-    int32_t line;
-
-    /** The column number. */
-    uint32_t column;
-} pm_line_column_t;
-
-/**
- * Initialize a new newline list with the given capacity. Returns true if the
- * allocation of the offsets succeeds, otherwise returns false.
- *
- * @param list The list to initialize.
- * @param start A pointer to the start of the source string.
- * @param capacity The initial capacity of the list.
- * @return True if the allocation of the offsets succeeds, otherwise false.
- */
-bool pm_newline_list_init(pm_newline_list_t *list, const uint8_t *start, size_t capacity);
-
-/**
- * Clear out the newlines that have been appended to the list.
- *
- * @param list The list to clear.
- */
-void
-pm_newline_list_clear(pm_newline_list_t *list);
-
-/**
- * Append a new offset to the newline list. Returns true if the reallocation of
- * the offsets succeeds (if one was necessary), otherwise returns false.
- *
- * @param list The list to append to.
- * @param cursor A pointer to the offset to append.
- * @return True if the reallocation of the offsets succeeds (if one was
- *     necessary), otherwise false.
- */
-bool pm_newline_list_append(pm_newline_list_t *list, const uint8_t *cursor);
-
-/**
- * Returns the line of the given offset. If the offset is not in the list, the
- * line of the closest offset less than the given offset is returned.
- *
- * @param list The list to search.
- * @param cursor A pointer to the offset to search for.
- * @param start_line The line to start counting from.
- * @return The line of the given offset.
- */
-int32_t pm_newline_list_line(const pm_newline_list_t *list, const uint8_t *cursor, int32_t start_line);
-
-/**
- * Returns the line and column of the given offset. If the offset is not in the
- * list, the line and column of the closest offset less than the given offset
- * are returned.
- *
- * @param list The list to search.
- * @param cursor A pointer to the offset to search for.
- * @param start_line The line to start counting from.
- * @return The line and column of the given offset.
- */
-pm_line_column_t pm_newline_list_line_column(const pm_newline_list_t *list, const uint8_t *cursor, int32_t start_line);
-
-/**
- * Free the internal memory allocated for the newline list.
- *
- * @param list The list to free.
- */
-void pm_newline_list_free(pm_newline_list_t *list);
-
-#endif

data/include/prism/util/pm_string.h

@@ -1,200 +0,0 @@
-/**
- * @file pm_string.h
- *
- * A generic string type that can have various ownership semantics.
- */
-#ifndef PRISM_STRING_H
-#define PRISM_STRING_H
-
-#include "prism/defines.h"
-
-#include <assert.h>
-#include <errno.h>
-#include <stdbool.h>
-#include <stddef.h>
-#include <stdlib.h>
-#include <string.h>
-
-// The following headers are necessary to read files using demand paging.
-#ifdef _WIN32
-#include <windows.h>
-#elif defined(_POSIX_MAPPED_FILES)
-#include <fcntl.h>
-#include <sys/mman.h>
-#include <sys/stat.h>
-#elif defined(PRISM_HAS_FILESYSTEM)
-#include <fcntl.h>
-#include <sys/stat.h>
-#endif
-
-/**
- * A generic string type that can have various ownership semantics.
- */
-typedef struct {
-    /** A pointer to the start of the string. */
-    const uint8_t *source;
-
-    /** The length of the string in bytes of memory. */
-    size_t length;
-
-    /** The type of the string. This field determines how the string should be freed. */
-    enum {
-        /** This string is a constant string, and should not be freed. */
-        PM_STRING_CONSTANT,
-
-        /** This is a slice of another string, and should not be freed. */
-        PM_STRING_SHARED,
-
-        /** This string owns its memory, and should be freed using `pm_string_free()`. */
-        PM_STRING_OWNED,
-
-#ifdef PRISM_HAS_MMAP
-        /** This string is a memory-mapped file, and should be freed using `pm_string_free()`. */
-        PM_STRING_MAPPED
-#endif
-    } type;
-} pm_string_t;
-
-/**
- * Returns the size of the pm_string_t struct. This is necessary to allocate the
- * correct amount of memory in the FFI backend.
- *
- * @return The size of the pm_string_t struct.
- */
-PRISM_EXPORTED_FUNCTION size_t pm_string_sizeof(void);
-
-/**
- * Defines an empty string. This is useful for initializing a string that will
- * be filled in later.
- */
-#define PM_STRING_EMPTY ((pm_string_t) { .type = PM_STRING_CONSTANT, .source = NULL, .length = 0 })
-
-/**
- * Initialize a shared string that is based on initial input.
- *
- * @param string The string to initialize.
- * @param start The start of the string.
- * @param end The end of the string.
- */
-void pm_string_shared_init(pm_string_t *string, const uint8_t *start, const uint8_t *end);
-
-/**
- * Initialize an owned string that is responsible for freeing allocated memory.
- *
- * @param string The string to initialize.
- * @param source The source of the string.
- * @param length The length of the string.
- */
-void pm_string_owned_init(pm_string_t *string, uint8_t *source, size_t length);
-
-/**
- * Initialize a constant string that doesn't own its memory source.
- *
- * @param string The string to initialize.
- * @param source The source of the string.
- * @param length The length of the string.
- */
-void pm_string_constant_init(pm_string_t *string, const char *source, size_t length);
-
-/**
- * Represents the result of calling pm_string_mapped_init or
- * pm_string_file_init. We need this additional information because there is
- * not a platform-agnostic way to indicate that the file that was attempted to
- * be opened was a directory.
- */
-typedef enum {
-    /** Indicates that the string was successfully initialized. */
-    PM_STRING_INIT_SUCCESS = 0,
-    /**
-     * Indicates a generic error from a string_*_init function, where the type
-     * of error should be read from `errno` or `GetLastError()`.
-     */
-    PM_STRING_INIT_ERROR_GENERIC = 1,
-    /**
-     * Indicates that the file that was attempted to be opened was a directory.
-     */
-    PM_STRING_INIT_ERROR_DIRECTORY = 2
-} pm_string_init_result_t;
-
-/**
- * Read the file indicated by the filepath parameter into source and load its
- * contents and size into the given `pm_string_t`. The given `pm_string_t`
- * should be freed using `pm_string_free` when it is no longer used.
- *
- * We want to use demand paging as much as possible in order to avoid having to
- * read the entire file into memory (which could be detrimental to performance
- * for large files). This means that if we're on windows we'll use
- * `MapViewOfFile`, on POSIX systems that have access to `mmap` we'll use
- * `mmap`, and on other POSIX systems we'll use `read`.
- *
- * @param string The string to initialize.
- * @param filepath The filepath to read.
- * @return The success of the read, indicated by the value of the enum.
- *
- * \public \memberof pm_string_t
- */
-PRISM_EXPORTED_FUNCTION pm_string_init_result_t pm_string_mapped_init(pm_string_t *string, const char *filepath);
-
-/**
- * Read the file indicated by the filepath parameter into source and load its
- * contents and size into the given `pm_string_t`. The given `pm_string_t`
- * should be freed using `pm_string_free` when it is no longer used.
- *
- * @param string The string to initialize.
- * @param filepath The filepath to read.
- * @return The success of the read, indicated by the value of the enum.
- *
- * \public \memberof pm_string_t
- */
-PRISM_EXPORTED_FUNCTION pm_string_init_result_t pm_string_file_init(pm_string_t *string, const char *filepath);
-
-/**
- * Ensure the string is owned. If it is not, then reinitialize it as owned and
- * copy over the previous source.
- *
- * @param string The string to ensure is owned.
- */
-void pm_string_ensure_owned(pm_string_t *string);
-
-/**
- * Compare the underlying lengths and bytes of two strings. Returns 0 if the
- * strings are equal, a negative number if the left string is less than the
- * right string, and a positive number if the left string is greater than the
- * right string.
- *
- * @param left The left string to compare.
- * @param right The right string to compare.
- * @return The comparison result.
- */
-int pm_string_compare(const pm_string_t *left, const pm_string_t *right);
-
-/**
- * Returns the length associated with the string.
- *
- * @param string The string to get the length of.
- * @return The length of the string.
- *
- * \public \memberof pm_string_t
- */
-PRISM_EXPORTED_FUNCTION size_t pm_string_length(const pm_string_t *string);
-
-/**
- * Returns the start pointer associated with the string.
- *
- * @param string The string to get the start pointer of.
- * @return The start pointer of the string.
- *
- * \public \memberof pm_string_t
- */
-PRISM_EXPORTED_FUNCTION const uint8_t * pm_string_source(const pm_string_t *string);
-
-/**
- * Free the associated memory of the given string.
- *
- * @param string The string to free.
- *
- * \public \memberof pm_string_t
- */
-PRISM_EXPORTED_FUNCTION void pm_string_free(pm_string_t *string);
-
-#endif

data/lib/prism/pack.rb

@@ -1,230 +0,0 @@
-# frozen_string_literal: true
-# :markup: markdown
-# typed: ignore
-
-#
-module Prism
-  # A parser for the pack template language.
-  module Pack
-    %i[
-      SPACE
-      COMMENT
-      INTEGER
-      UTF8
-      BER
-      FLOAT
-      STRING_SPACE_PADDED
-      STRING_NULL_PADDED
-      STRING_NULL_TERMINATED
-      STRING_MSB
-      STRING_LSB
-      STRING_HEX_HIGH
-      STRING_HEX_LOW
-      STRING_UU
-      STRING_MIME
-      STRING_BASE64
-      STRING_FIXED
-      STRING_POINTER
-      MOVE
-      BACK
-      NULL
-
-      UNSIGNED
-      SIGNED
-      SIGNED_NA
-
-      AGNOSTIC_ENDIAN
-      LITTLE_ENDIAN
-      BIG_ENDIAN
-      NATIVE_ENDIAN
-      ENDIAN_NA
-
-      SIZE_SHORT
-      SIZE_INT
-      SIZE_LONG
-      SIZE_LONG_LONG
-      SIZE_8
-      SIZE_16
-      SIZE_32
-      SIZE_64
-      SIZE_P
-      SIZE_NA
-
-      LENGTH_FIXED
-      LENGTH_MAX
-      LENGTH_RELATIVE
-      LENGTH_NA
-    ].each do |const|
-      const_set(const, const)
-    end
-
-    # A directive in the pack template language.
-    class Directive
-      # A symbol representing the version of Ruby.
-      attr_reader :version
-
-      # A symbol representing whether or not we are packing or unpacking.
-      attr_reader :variant
-
-      # A byteslice of the source string that this directive represents.
-      attr_reader :source
-
-      # The type of the directive.
-      attr_reader :type
-
-      # The type of signedness of the directive.
-      attr_reader :signed
-
-      # The type of endianness of the directive.
-      attr_reader :endian
-
-      # The size of the directive.
-      attr_reader :size
-
-      # The length type of this directive (used for integers).
-      attr_reader :length_type
-
-      # The length of this directive (used for integers).
-      attr_reader :length
-
-      # Initialize a new directive with the given values.
-      def initialize(version, variant, source, type, signed, endian, size, length_type, length)
-        @version = version
-        @variant = variant
-        @source = source
-        @type = type
-        @signed = signed
-        @endian = endian
-        @size = size
-        @length_type = length_type
-        @length = length
-      end
-
-      # The descriptions of the various types of endianness.
-      ENDIAN_DESCRIPTIONS = {
-        AGNOSTIC_ENDIAN: "agnostic",
-        LITTLE_ENDIAN: "little-endian (VAX)",
-        BIG_ENDIAN: "big-endian (network)",
-        NATIVE_ENDIAN: "native-endian",
-        ENDIAN_NA: "n/a"
-      }
-
-      # The descriptions of the various types of signedness.
-      SIGNED_DESCRIPTIONS = {
-        UNSIGNED: "unsigned",
-        SIGNED: "signed",
-        SIGNED_NA: "n/a"
-      }
-
-      # The descriptions of the various types of sizes.
-      SIZE_DESCRIPTIONS = {
-        SIZE_SHORT: "short",
-        SIZE_INT: "int-width",
-        SIZE_LONG: "long",
-        SIZE_LONG_LONG: "long long",
-        SIZE_8: "8-bit",
-        SIZE_16: "16-bit",
-        SIZE_32: "32-bit",
-        SIZE_64: "64-bit",
-        SIZE_P: "pointer-width"
-      }
-
-      # Provide a human-readable description of the directive.
-      def describe
-        case type
-        when SPACE
-          "whitespace"
-        when COMMENT
-          "comment"
-        when INTEGER
-          if size == SIZE_8
-            base = "#{SIGNED_DESCRIPTIONS[signed]} #{SIZE_DESCRIPTIONS[size]} integer"
-          else
-            base = "#{SIGNED_DESCRIPTIONS[signed]} #{SIZE_DESCRIPTIONS[size]} #{ENDIAN_DESCRIPTIONS[endian]} integer"
-          end
-          case length_type
-          when LENGTH_FIXED
-            if length > 1
-              base + ", x#{length}"
-            else
-              base
-            end
-          when LENGTH_MAX
-            base + ", as many as possible"
-          else
-            raise
-          end
-        when UTF8
-          "UTF-8 character"
-        when BER
-          "BER-compressed integer"
-        when FLOAT
-          "#{SIZE_DESCRIPTIONS[size]} #{ENDIAN_DESCRIPTIONS[endian]} float"
-        when STRING_SPACE_PADDED
-          "arbitrary binary string (space padded)"
-        when STRING_NULL_PADDED
-          "arbitrary binary string (null padded, count is width)"
-        when STRING_NULL_TERMINATED
-          "arbitrary binary string (null padded, count is width), except that null is added with *"
-        when STRING_MSB
-          "bit string (MSB first)"
-        when STRING_LSB
-          "bit string (LSB first)"
-        when STRING_HEX_HIGH
-          "hex string (high nibble first)"
-        when STRING_HEX_LOW
-          "hex string (low nibble first)"
-        when STRING_UU
-          "UU-encoded string"
-        when STRING_MIME
-          "quoted printable, MIME encoding"
-        when STRING_BASE64
-          "base64 encoded string"
-        when STRING_FIXED
-          "pointer to a structure (fixed-length string)"
-        when STRING_POINTER
-          "pointer to a null-terminated string"
-        when MOVE
-          "move to absolute position"
-        when BACK
-          "back up a byte"
-        when NULL
-          "null byte"
-        else
-          raise
-        end
-      end
-    end
-
-    # The result of parsing a pack template.
-    class Format
-      # A list of the directives in the template.
-      attr_reader :directives
-
-      # The encoding of the template.
-      attr_reader :encoding
-
-      # Create a new Format with the given directives and encoding.
-      def initialize(directives, encoding)
-        @directives = directives
-        @encoding = encoding
-      end
-
-      # Provide a human-readable description of the format.
-      def describe
-        source_width = directives.map { |d| d.source.inspect.length }.max
-        directive_lines = directives.map do |directive|
-          if directive.type == SPACE
-            source = directive.source.inspect
-          else
-            source = directive.source
-          end
-          # @type var source_width: Integer
-          "  #{source.ljust(source_width)}  #{directive.describe}"
-        end
-
-        (["Directives:"] + directive_lines + ["Encoding:", "  #{encoding}"]).join("\n")
-      end
-    end
-  end
-end