prism 0.13.0

data/include/prism/defines.h

@@ -0,0 +1,45 @@
+#ifndef PRISM_DEFINES_H
+#define PRISM_DEFINES_H
+
+// This file should be included first by any *.h or *.c in prism
+
+#include <ctype.h>
+#include <stdarg.h>
+#include <stddef.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <string.h>
+
+// PRISM_EXPORTED_FUNCTION
+#ifndef PRISM_EXPORTED_FUNCTION
+#   ifdef PRISM_EXPORT_SYMBOLS
+#       ifdef _WIN32
+#          define PRISM_EXPORTED_FUNCTION __declspec(dllexport) extern
+#       else
+#          define PRISM_EXPORTED_FUNCTION __attribute__((__visibility__("default"))) extern
+#       endif
+#   else
+#       define PRISM_EXPORTED_FUNCTION
+#   endif
+#endif
+
+// PRISM_ATTRIBUTE_UNUSED
+#if defined(__GNUC__)
+#   define PRISM_ATTRIBUTE_UNUSED __attribute__((unused))
+#else
+#   define PRISM_ATTRIBUTE_UNUSED
+#endif
+
+// inline
+#if defined(_MSC_VER) && !defined(inline)
+#   define inline __inline
+#endif
+
+// Windows versions before 2015 use _snprintf
+#if !defined(snprintf) && defined(_MSC_VER) && (_MSC_VER < 1900)
+#   define snprintf _snprintf
+#endif
+
+int pm_strncasecmp(const uint8_t *string1, const uint8_t *string2, size_t length);
+
+#endif

data/include/prism/diagnostic.h

@@ -0,0 +1,231 @@
+#ifndef PRISM_DIAGNOSTIC_H
+#define PRISM_DIAGNOSTIC_H
+
+#include "prism/defines.h"
+#include "prism/util/pm_list.h"
+
+#include <stdbool.h>
+#include <stdlib.h>
+#include <assert.h>
+
+// This struct represents a diagnostic found during parsing.
+typedef struct {
+    pm_list_node_t node;
+    const uint8_t *start;
+    const uint8_t *end;
+    const char *message;
+} pm_diagnostic_t;
+
+typedef enum {
+    PM_ERR_ALIAS_ARGUMENT,
+    PM_ERR_AMPAMPEQ_MULTI_ASSIGN,
+    PM_ERR_ARGUMENT_AFTER_BLOCK,
+    PM_ERR_ARGUMENT_BARE_HASH,
+    PM_ERR_ARGUMENT_BLOCK_MULTI,
+    PM_ERR_ARGUMENT_FORMAL_CLASS,
+    PM_ERR_ARGUMENT_FORMAL_CONSTANT,
+    PM_ERR_ARGUMENT_FORMAL_GLOBAL,
+    PM_ERR_ARGUMENT_FORMAL_IVAR,
+    PM_ERR_ARGUMENT_NO_FORWARDING_AMP,
+    PM_ERR_ARGUMENT_NO_FORWARDING_ELLIPSES,
+    PM_ERR_ARGUMENT_NO_FORWARDING_STAR,
+    PM_ERR_ARGUMENT_SPLAT_AFTER_ASSOC_SPLAT,
+    PM_ERR_ARGUMENT_SPLAT_AFTER_SPLAT,
+    PM_ERR_ARGUMENT_TERM_PAREN,
+    PM_ERR_ARGUMENT_UNEXPECTED_BLOCK,
+    PM_ERR_ARRAY_ELEMENT,
+    PM_ERR_ARRAY_EXPRESSION,
+    PM_ERR_ARRAY_EXPRESSION_AFTER_STAR,
+    PM_ERR_ARRAY_SEPARATOR,
+    PM_ERR_ARRAY_TERM,
+    PM_ERR_BEGIN_LONELY_ELSE,
+    PM_ERR_BEGIN_TERM,
+    PM_ERR_BEGIN_UPCASE_BRACE,
+    PM_ERR_BEGIN_UPCASE_TERM,
+    PM_ERR_BEGIN_UPCASE_TOPLEVEL,
+    PM_ERR_BLOCK_PARAM_LOCAL_VARIABLE,
+    PM_ERR_BLOCK_PARAM_PIPE_TERM,
+    PM_ERR_BLOCK_TERM_BRACE,
+    PM_ERR_BLOCK_TERM_END,
+    PM_ERR_CANNOT_PARSE_EXPRESSION,
+    PM_ERR_CANNOT_PARSE_STRING_PART,
+    PM_ERR_CASE_EXPRESSION_AFTER_CASE,
+    PM_ERR_CASE_EXPRESSION_AFTER_WHEN,
+    PM_ERR_CASE_MISSING_CONDITIONS,
+    PM_ERR_CASE_TERM,
+    PM_ERR_CLASS_IN_METHOD,
+    PM_ERR_CLASS_NAME,
+    PM_ERR_CLASS_SUPERCLASS,
+    PM_ERR_CLASS_TERM,
+    PM_ERR_CLASS_UNEXPECTED_END,
+    PM_ERR_CONDITIONAL_ELSIF_PREDICATE,
+    PM_ERR_CONDITIONAL_IF_PREDICATE,
+    PM_ERR_CONDITIONAL_PREDICATE_TERM,
+    PM_ERR_CONDITIONAL_TERM,
+    PM_ERR_CONDITIONAL_TERM_ELSE,
+    PM_ERR_CONDITIONAL_UNLESS_PREDICATE,
+    PM_ERR_CONDITIONAL_UNTIL_PREDICATE,
+    PM_ERR_CONDITIONAL_WHILE_PREDICATE,
+    PM_ERR_CONSTANT_PATH_COLON_COLON_CONSTANT,
+    PM_ERR_DEF_ENDLESS,
+    PM_ERR_DEF_ENDLESS_SETTER,
+    PM_ERR_DEF_NAME,
+    PM_ERR_DEF_NAME_AFTER_RECEIVER,
+    PM_ERR_DEF_PARAMS_TERM,
+    PM_ERR_DEF_PARAMS_TERM_PAREN,
+    PM_ERR_DEF_RECEIVER,
+    PM_ERR_DEF_RECEIVER_TERM,
+    PM_ERR_DEF_TERM,
+    PM_ERR_DEFINED_EXPRESSION,
+    PM_ERR_EMBDOC_TERM,
+    PM_ERR_EMBEXPR_END,
+    PM_ERR_EMBVAR_INVALID,
+    PM_ERR_END_UPCASE_BRACE,
+    PM_ERR_END_UPCASE_TERM,
+    PM_ERR_ESCAPE_INVALID_CONTROL,
+    PM_ERR_ESCAPE_INVALID_CONTROL_REPEAT,
+    PM_ERR_ESCAPE_INVALID_HEXADECIMAL,
+    PM_ERR_ESCAPE_INVALID_META,
+    PM_ERR_ESCAPE_INVALID_META_REPEAT,
+    PM_ERR_ESCAPE_INVALID_UNICODE,
+    PM_ERR_ESCAPE_INVALID_UNICODE_CM_FLAGS,
+    PM_ERR_ESCAPE_INVALID_UNICODE_LITERAL,
+    PM_ERR_ESCAPE_INVALID_UNICODE_LONG,
+    PM_ERR_ESCAPE_INVALID_UNICODE_TERM,
+    PM_ERR_EXPECT_ARGUMENT,
+    PM_ERR_EXPECT_EOL_AFTER_STATEMENT,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_AMPAMPEQ,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_PIPEPIPEEQ,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_COMMA,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_EQUAL,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_LESS_LESS,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_LPAREN,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_QUESTION,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_OPERATOR,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_SPLAT,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_SPLAT_HASH,
+    PM_ERR_EXPECT_EXPRESSION_AFTER_STAR,
+    PM_ERR_EXPECT_IDENT_REQ_PARAMETER,
+    PM_ERR_EXPECT_LPAREN_REQ_PARAMETER,
+    PM_ERR_EXPECT_RBRACKET,
+    PM_ERR_EXPECT_RPAREN,
+    PM_ERR_EXPECT_RPAREN_AFTER_MULTI,
+    PM_ERR_EXPECT_RPAREN_REQ_PARAMETER,
+    PM_ERR_EXPECT_STRING_CONTENT,
+    PM_ERR_EXPECT_WHEN_DELIMITER,
+    PM_ERR_EXPRESSION_BARE_HASH,
+    PM_ERR_FOR_COLLECTION,
+    PM_ERR_FOR_IN,
+    PM_ERR_FOR_INDEX,
+    PM_ERR_FOR_TERM,
+    PM_ERR_HASH_EXPRESSION_AFTER_LABEL,
+    PM_ERR_HASH_KEY,
+    PM_ERR_HASH_ROCKET,
+    PM_ERR_HASH_TERM,
+    PM_ERR_HASH_VALUE,
+    PM_ERR_HEREDOC_TERM,
+    PM_ERR_INCOMPLETE_QUESTION_MARK,
+    PM_ERR_INCOMPLETE_VARIABLE_CLASS,
+    PM_ERR_INCOMPLETE_VARIABLE_INSTANCE,
+    PM_ERR_INVALID_ENCODING_MAGIC_COMMENT,
+    PM_ERR_INVALID_FLOAT_EXPONENT,
+    PM_ERR_INVALID_NUMBER_BINARY,
+    PM_ERR_INVALID_NUMBER_DECIMAL,
+    PM_ERR_INVALID_NUMBER_HEXADECIMAL,
+    PM_ERR_INVALID_NUMBER_OCTAL,
+    PM_ERR_INVALID_NUMBER_UNDERSCORE,
+    PM_ERR_INVALID_PERCENT,
+    PM_ERR_INVALID_TOKEN,
+    PM_ERR_INVALID_VARIABLE_GLOBAL,
+    PM_ERR_LAMBDA_OPEN,
+    PM_ERR_LAMBDA_TERM_BRACE,
+    PM_ERR_LAMBDA_TERM_END,
+    PM_ERR_LIST_I_LOWER_ELEMENT,
+    PM_ERR_LIST_I_LOWER_TERM,
+    PM_ERR_LIST_I_UPPER_ELEMENT,
+    PM_ERR_LIST_I_UPPER_TERM,
+    PM_ERR_LIST_W_LOWER_ELEMENT,
+    PM_ERR_LIST_W_LOWER_TERM,
+    PM_ERR_LIST_W_UPPER_ELEMENT,
+    PM_ERR_LIST_W_UPPER_TERM,
+    PM_ERR_MALLOC_FAILED,
+    PM_ERR_MODULE_IN_METHOD,
+    PM_ERR_MODULE_NAME,
+    PM_ERR_MODULE_TERM,
+    PM_ERR_MULTI_ASSIGN_MULTI_SPLATS,
+    PM_ERR_NOT_EXPRESSION,
+    PM_ERR_NUMBER_LITERAL_UNDERSCORE,
+    PM_ERR_NUMBERED_PARAMETER_NOT_ALLOWED,
+    PM_ERR_NUMBERED_PARAMETER_OUTER_SCOPE,
+    PM_ERR_OPERATOR_MULTI_ASSIGN,
+    PM_ERR_OPERATOR_WRITE_BLOCK,
+    PM_ERR_PARAMETER_ASSOC_SPLAT_MULTI,
+    PM_ERR_PARAMETER_BLOCK_MULTI,
+    PM_ERR_PARAMETER_METHOD_NAME,
+    PM_ERR_PARAMETER_NAME_REPEAT,
+    PM_ERR_PARAMETER_NO_DEFAULT,
+    PM_ERR_PARAMETER_NO_DEFAULT_KW,
+    PM_ERR_PARAMETER_NUMBERED_RESERVED,
+    PM_ERR_PARAMETER_ORDER,
+    PM_ERR_PARAMETER_SPLAT_MULTI,
+    PM_ERR_PARAMETER_STAR,
+    PM_ERR_PARAMETER_UNEXPECTED_FWD,
+    PM_ERR_PARAMETER_WILD_LOOSE_COMMA,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_BRACKET,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_HROCKET,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_COMMA,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_IN,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_KEY,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_PAREN,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_PIN,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_PIPE,
+    PM_ERR_PATTERN_EXPRESSION_AFTER_RANGE,
+    PM_ERR_PATTERN_HASH_KEY,
+    PM_ERR_PATTERN_HASH_KEY_LABEL,
+    PM_ERR_PATTERN_IDENT_AFTER_HROCKET,
+    PM_ERR_PATTERN_LABEL_AFTER_COMMA,
+    PM_ERR_PATTERN_REST,
+    PM_ERR_PATTERN_TERM_BRACE,
+    PM_ERR_PATTERN_TERM_BRACKET,
+    PM_ERR_PATTERN_TERM_PAREN,
+    PM_ERR_PIPEPIPEEQ_MULTI_ASSIGN,
+    PM_ERR_REGEXP_TERM,
+    PM_ERR_RESCUE_EXPRESSION,
+    PM_ERR_RESCUE_MODIFIER_VALUE,
+    PM_ERR_RESCUE_TERM,
+    PM_ERR_RESCUE_VARIABLE,
+    PM_ERR_RETURN_INVALID,
+    PM_ERR_STRING_CONCATENATION,
+    PM_ERR_STRING_INTERPOLATED_TERM,
+    PM_ERR_STRING_LITERAL_TERM,
+    PM_ERR_SYMBOL_INVALID,
+    PM_ERR_SYMBOL_TERM_DYNAMIC,
+    PM_ERR_SYMBOL_TERM_INTERPOLATED,
+    PM_ERR_TERNARY_COLON,
+    PM_ERR_TERNARY_EXPRESSION_FALSE,
+    PM_ERR_TERNARY_EXPRESSION_TRUE,
+    PM_ERR_UNARY_RECEIVER_BANG,
+    PM_ERR_UNARY_RECEIVER_MINUS,
+    PM_ERR_UNARY_RECEIVER_PLUS,
+    PM_ERR_UNARY_RECEIVER_TILDE,
+    PM_ERR_UNDEF_ARGUMENT,
+    PM_ERR_UNTIL_TERM,
+    PM_ERR_WHILE_TERM,
+    PM_ERR_WRITE_TARGET_READONLY,
+    PM_ERR_WRITE_TARGET_UNEXPECTED,
+    PM_ERR_XSTRING_TERM,
+    PM_WARN_AMBIGUOUS_FIRST_ARGUMENT_MINUS,
+    PM_WARN_AMBIGUOUS_FIRST_ARGUMENT_PLUS,
+    PM_WARN_AMBIGUOUS_PREFIX_STAR,
+    PM_WARN_AMBIGUOUS_SLASH,
+    /* This must be the last member. */
+    PM_DIAGNOSTIC_ID_LEN,
+} pm_diagnostic_id_t;
+
+// Append a diagnostic to the given list of diagnostics.
+bool pm_diagnostic_list_append(pm_list_t *list, const uint8_t *start, const uint8_t *end, pm_diagnostic_id_t diag_id);
+
+// Deallocate the internal state of the given diagnostic list.
+void pm_diagnostic_list_free(pm_list_t *list);
+
+#endif

data/include/prism/enc/pm_encoding.h

@@ -0,0 +1,95 @@
+#ifndef PRISM_ENCODING_H
+#define PRISM_ENCODING_H
+
+#include "prism/defines.h"
+
+#include <assert.h>
+#include <stdbool.h>
+#include <stddef.h>
+#include <stdint.h>
+
+// This struct defines the functions necessary to implement the encoding
+// interface so we can determine how many bytes the subsequent character takes.
+// Each callback should return the number of bytes, or 0 if the next bytes are
+// invalid for the encoding and type.
+typedef struct {
+    // Return the number of bytes that the next character takes if it is valid
+    // in the encoding. Does not read more than n bytes. It is assumed that n is
+    // at least 1.
+    size_t (*char_width)(const uint8_t *b, ptrdiff_t n);
+
+    // Return the number of bytes that the next character takes if it is valid
+    // in the encoding and is alphabetical. Does not read more than n bytes. It
+    // is assumed that n is at least 1.
+    size_t (*alpha_char)(const uint8_t *b, ptrdiff_t n);
+
+    // Return the number of bytes that the next character takes if it is valid
+    // in the encoding and is alphanumeric. Does not read more than n bytes. It
+    // is assumed that n is at least 1.
+    size_t (*alnum_char)(const uint8_t *b, ptrdiff_t n);
+
+    // Return true if the next character is valid in the encoding and is an
+    // uppercase character. Does not read more than n bytes. It is assumed that
+    // n is at least 1.
+    bool (*isupper_char)(const uint8_t *b, ptrdiff_t n);
+
+    // The name of the encoding. This should correspond to a value that can be
+    // passed to Encoding.find in Ruby.
+    const char *name;
+
+    // Return true if the encoding is a multibyte encoding.
+    bool multibyte;
+} pm_encoding_t;
+
+// These bits define the location of each bit of metadata within the various
+// lookup tables that are used to determine the properties of a character.
+#define PRISM_ENCODING_ALPHABETIC_BIT 1 << 0
+#define PRISM_ENCODING_ALPHANUMERIC_BIT 1 << 1
+#define PRISM_ENCODING_UPPERCASE_BIT 1 << 2
+
+// These functions are reused by some other encodings, so they are defined here
+// so they can be shared.
+size_t pm_encoding_ascii_alpha_char(const uint8_t *b, PRISM_ATTRIBUTE_UNUSED ptrdiff_t n);
+size_t pm_encoding_ascii_alnum_char(const uint8_t *b, PRISM_ATTRIBUTE_UNUSED ptrdiff_t n);
+bool pm_encoding_ascii_isupper_char(const uint8_t *b, PRISM_ATTRIBUTE_UNUSED ptrdiff_t n);
+
+// These functions are shared between the actual encoding and the fast path in
+// the parser so they need to be internally visible.
+size_t pm_encoding_utf_8_alpha_char(const uint8_t *b, ptrdiff_t n);
+size_t pm_encoding_utf_8_alnum_char(const uint8_t *b, ptrdiff_t n);
+
+// This lookup table is referenced in both the UTF-8 encoding file and the
+// parser directly in order to speed up the default encoding processing.
+extern const uint8_t pm_encoding_unicode_table[256];
+
+// These are the encodings that are supported by the parser. They are defined in
+// their own files in the src/enc directory.
+extern pm_encoding_t pm_encoding_ascii;
+extern pm_encoding_t pm_encoding_ascii_8bit;
+extern pm_encoding_t pm_encoding_big5;
+extern pm_encoding_t pm_encoding_euc_jp;
+extern pm_encoding_t pm_encoding_gbk;
+extern pm_encoding_t pm_encoding_iso_8859_1;
+extern pm_encoding_t pm_encoding_iso_8859_2;
+extern pm_encoding_t pm_encoding_iso_8859_3;
+extern pm_encoding_t pm_encoding_iso_8859_4;
+extern pm_encoding_t pm_encoding_iso_8859_5;
+extern pm_encoding_t pm_encoding_iso_8859_6;
+extern pm_encoding_t pm_encoding_iso_8859_7;
+extern pm_encoding_t pm_encoding_iso_8859_8;
+extern pm_encoding_t pm_encoding_iso_8859_9;
+extern pm_encoding_t pm_encoding_iso_8859_10;
+extern pm_encoding_t pm_encoding_iso_8859_11;
+extern pm_encoding_t pm_encoding_iso_8859_13;
+extern pm_encoding_t pm_encoding_iso_8859_14;
+extern pm_encoding_t pm_encoding_iso_8859_15;
+extern pm_encoding_t pm_encoding_iso_8859_16;
+extern pm_encoding_t pm_encoding_koi8_r;
+extern pm_encoding_t pm_encoding_shift_jis;
+extern pm_encoding_t pm_encoding_utf_8;
+extern pm_encoding_t pm_encoding_utf8_mac;
+extern pm_encoding_t pm_encoding_windows_31j;
+extern pm_encoding_t pm_encoding_windows_1251;
+extern pm_encoding_t pm_encoding_windows_1252;
+
+#endif

data/include/prism/node.h

@@ -0,0 +1,41 @@
+#ifndef PRISM_NODE_H
+#define PRISM_NODE_H
+
+#include "prism/defines.h"
+#include "prism/parser.h"
+
+// Append a new node onto the end of the node list.
+void pm_node_list_append(pm_node_list_t *list, pm_node_t *node);
+
+// Clear the node but preserves the location.
+void pm_node_clear(pm_node_t *node);
+
+// Deallocate a node and all of its children.
+PRISM_EXPORTED_FUNCTION void pm_node_destroy(pm_parser_t *parser, struct pm_node *node);
+
+// This struct stores the information gathered by the pm_node_memsize function.
+// It contains both the memory footprint and additionally metadata about the
+// shape of the tree.
+typedef struct {
+    size_t memsize;
+    size_t node_count;
+} pm_memsize_t;
+
+// Calculates the memory footprint of a given node.
+PRISM_EXPORTED_FUNCTION void pm_node_memsize(pm_node_t *node, pm_memsize_t *memsize);
+
+// Returns a string representation of the given node type.
+PRISM_EXPORTED_FUNCTION const char * pm_node_type_to_str(pm_node_type_t node_type);
+
+#define PM_EMPTY_NODE_LIST ((pm_node_list_t) { .nodes = NULL, .size = 0, .capacity = 0 })
+
+// ScopeNodes are helper nodes, and will never be part of the AST. We manually
+// declare them here to avoid generating them.
+typedef struct pm_scope_node {
+    pm_node_t base;
+    struct pm_parameters_node *parameters;
+    pm_node_t *body;
+    pm_constant_id_list_t locals;
+} pm_scope_node_t;
+
+#endif // PRISM_NODE_H

data/include/prism/pack.h

@@ -0,0 +1,141 @@
+#ifndef PRISM_PACK_H
+#define PRISM_PACK_H
+
+#include "prism/defines.h"
+
+#include <stdint.h>
+#include <stdlib.h>
+
+typedef enum pm_pack_version {
+    PM_PACK_VERSION_3_2_0
+} pm_pack_version;
+
+typedef enum pm_pack_variant {
+    PM_PACK_VARIANT_PACK,
+    PM_PACK_VARIANT_UNPACK
+} pm_pack_variant;
+
+typedef enum pm_pack_type {
+    PM_PACK_SPACE,
+    PM_PACK_COMMENT,
+    PM_PACK_INTEGER,
+    PM_PACK_UTF8,
+    PM_PACK_BER,
+    PM_PACK_FLOAT,
+    PM_PACK_STRING_SPACE_PADDED,
+    PM_PACK_STRING_NULL_PADDED,
+    PM_PACK_STRING_NULL_TERMINATED,
+    PM_PACK_STRING_MSB,
+    PM_PACK_STRING_LSB,
+    PM_PACK_STRING_HEX_HIGH,
+    PM_PACK_STRING_HEX_LOW,
+    PM_PACK_STRING_UU,
+    PM_PACK_STRING_MIME,
+    PM_PACK_STRING_BASE64,
+    PM_PACK_STRING_FIXED,
+    PM_PACK_STRING_POINTER,
+    PM_PACK_MOVE,
+    PM_PACK_BACK,
+    PM_PACK_NULL,
+    PM_PACK_END
+} pm_pack_type;
+
+typedef enum pm_pack_signed {
+    PM_PACK_UNSIGNED,
+    PM_PACK_SIGNED,
+    PM_PACK_SIGNED_NA
+} pm_pack_signed;
+
+typedef enum pm_pack_endian {
+    PM_PACK_AGNOSTIC_ENDIAN,
+    PM_PACK_LITTLE_ENDIAN,      // aka 'VAX', or 'V'
+    PM_PACK_BIG_ENDIAN,         // aka 'network', or 'N'
+    PM_PACK_NATIVE_ENDIAN,
+    PM_PACK_ENDIAN_NA
+} pm_pack_endian;
+
+typedef enum pm_pack_size {
+    PM_PACK_SIZE_SHORT,
+    PM_PACK_SIZE_INT,
+    PM_PACK_SIZE_LONG,
+    PM_PACK_SIZE_LONG_LONG,
+    PM_PACK_SIZE_8,
+    PM_PACK_SIZE_16,
+    PM_PACK_SIZE_32,
+    PM_PACK_SIZE_64,
+    PM_PACK_SIZE_P,
+    PM_PACK_SIZE_NA
+} pm_pack_size;
+
+typedef enum pm_pack_length_type {
+    PM_PACK_LENGTH_FIXED,
+    PM_PACK_LENGTH_MAX,
+    PM_PACK_LENGTH_RELATIVE,  // special case for unpack @*
+    PM_PACK_LENGTH_NA
+} pm_pack_length_type;
+
+typedef enum pm_pack_encoding {
+    PM_PACK_ENCODING_START,
+    PM_PACK_ENCODING_ASCII_8BIT,
+    PM_PACK_ENCODING_US_ASCII,
+    PM_PACK_ENCODING_UTF_8
+} pm_pack_encoding;
+
+typedef enum pm_pack_result {
+    PM_PACK_OK,
+    PM_PACK_ERROR_UNSUPPORTED_DIRECTIVE,
+    PM_PACK_ERROR_UNKNOWN_DIRECTIVE,
+    PM_PACK_ERROR_LENGTH_TOO_BIG,
+    PM_PACK_ERROR_BANG_NOT_ALLOWED,
+    PM_PACK_ERROR_DOUBLE_ENDIAN
+} pm_pack_result;
+
+// Parse a single directive from a pack or unpack format string.
+//
+// Parameters:
+//  - [in] pm_pack_version version    the version of Ruby
+//  - [in] pm_pack_variant variant    pack or unpack
+//  - [in out] const char **format    the start of the next directive to parse
+//      on calling, and advanced beyond the parsed directive on return, or as
+//      much of it as was consumed until an error was encountered
+//  - [in] const char *format_end     the end of the format string
+//  - [out] pm_pack_type *type        the type of the directive
+//  - [out] pm_pack_signed *signed_type
+//                                    whether the value is signed
+//  - [out] pm_pack_endian *endian    the endianness of the value
+//  - [out] pm_pack_size *size        the size of the value
+//  - [out] pm_pack_length_type *length_type
+//                                    what kind of length is specified
+//  - [out] size_t *length            the length of the directive
+//  - [in out] pm_pack_encoding *encoding
+//                                    takes the current encoding of the string
+//      which would result from parsing the whole format string, and returns a
+//      possibly changed directive - the encoding should be
+//      PM_PACK_ENCODING_START when pm_pack_parse is called for the first
+//      directive in a format string
+//
+// Return:
+//  - PM_PACK_OK on success
+//  - PM_PACK_ERROR_* on error
+//
+// Notes:
+//   Consult Ruby documentation for the meaning of directives.
+PRISM_EXPORTED_FUNCTION pm_pack_result
+pm_pack_parse(
+    pm_pack_variant variant_arg,
+    const char **format,
+    const char *format_end,
+    pm_pack_type *type,
+    pm_pack_signed *signed_type,
+    pm_pack_endian *endian,
+    pm_pack_size *size,
+    pm_pack_length_type *length_type,
+    uint64_t *length,
+    pm_pack_encoding *encoding
+);
+
+// prism abstracts sizes away from the native system - this converts an abstract
+// size to a native size.
+PRISM_EXPORTED_FUNCTION size_t pm_size_to_native(pm_pack_size size);
+
+#endif