src/parse.c

/**
 * @file parse.c
 * @author Michal Vasko <mvasko@cesnet.cz>
 * @brief parsing functions implementation
 *
 * Copyright (c) 2015 CESNET, z.s.p.o.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in
 *    the documentation and/or other materials provided with the
 *    distribution.
 * 3. Neither the name of the Company nor the names of its contributors
 *    may be used to endorse or promote products derived from this
 *    software without specific prior written permission.
 */

#include <ctype.h>
#include <assert.h>
#include <string.h>
#include <stdlib.h>
#include <errno.h>

#include "parse.h"
#include "resolve.h"
#include "common.h"

/**
 * @brief Checks the syntax of length or range statement,
 *        on success checks the semantics as well.
 *
 * @param[in] expr Length or range expression.
 * @param[in] type Type with the restriction.
 *
 * @return EXIT_SUCCESS on success, EXIT_FAILURE otherwise.
 */
int
check_length_range(const char *expr, struct lys_type *type)
{
    struct len_ran_intv *intv = NULL, *tmp_intv;
    const char *c = expr;
    char *tail;
    int ret = EXIT_FAILURE, flg = 1; /* first run flag */

    assert(expr);

lengthpart:

    while (isspace(*c)) {
        c++;
    }

    /* lower boundary or explicit number */
    if (!strncmp(c, "max", 3)) {
max:
        c += 3;
        while (isspace(*c)) {
            c++;
        }
        if (*c != '\0') {
            goto error;
        }

        goto syntax_ok;

    } else if (!strncmp(c, "min", 3)) {
        if (!flg) {
            /* min cannot be used elsewhere than in the first length-part */
            goto error;
        } else {
            flg = 0;
        }
        c += 3;
        while (isspace(*c)) {
            c++;
        }

        if (*c == '|') {
            c++;
            /* process next length-parth */
            goto lengthpart;
        } else if (*c == '\0') {
            goto syntax_ok;
        } else if (!strncmp(c, "..", 2)) {
upper:
            c += 2;
            while (isspace(*c)) {
                c++;
            }
            if (*c == '\0') {
                goto error;
            }

            /* upper boundary */
            if (!strncmp(c, "max", 3)) {
                goto max;
            }

            if (!isdigit(*c) && (*c != '+') && (*c != '-')) {
                goto error;
            }

            errno = 0;
            strtoll(c, &tail, 10);
            if (errno) {
                goto error;
            }
            c = tail;
            while (isspace(*c)) {
                c++;
            }
            if (*c == '\0') {
                goto syntax_ok;
            } else if (*c == '|') {
                c++;
                /* process next length-parth */
                goto lengthpart;
            } else {
                goto error;
            }
        } else {
            goto error;
        }

    } else if (isdigit(*c) || (*c == '-') || (*c == '+')) {
        /* number */
        errno = 0;
        strtoll(c, &tail, 10);
        if (errno) {
            /* out of range value */
            goto error;
        }
        c = tail;
        while (isspace(*c)) {
            c++;
        }

        if (*c == '|') {
            c++;
            /* process next length-parth */
            goto lengthpart;
        } else if (*c == '\0') {
            goto syntax_ok;
        } else if (!strncmp(c, "..", 2)) {
            goto upper;
        }

    } else {
        goto error;
    }

syntax_ok:

    if (resolve_len_ran_interval(expr, type, 1, &intv)) {
        goto error;
    }

    ret = EXIT_SUCCESS;

error:

    while (intv) {
        tmp_intv = intv->next;
        free(intv);
        intv = tmp_intv;
    }

    return ret;
}

/**
 * @brief Parse an identifier.
 *
 * ;; An identifier MUST NOT start with (('X'|'x') ('M'|'m') ('L'|'l'))
 * identifier          = (ALPHA / "_")
 *                       *(ALPHA / DIGIT / "_" / "-" / ".")
 *
 * @param[in] id Identifier in question.
 *
 * @return Number of characters successfully parsed.
 */
static int
parse_identifier(const char *id)
{
    int parsed = 0;

    if (((id[0] == 'x') || (id[0] == 'X'))
            && ((id[1] == 'm') || (id[0] == 'M'))
            && ((id[2] == 'l') || (id[2] == 'L'))) {
        return -parsed;
    }

    if (!isalpha(id[0]) && (id[0] != '_')) {
        return -parsed;
    }

    ++parsed;
    ++id;

    while (isalnum(id[0]) || (id[0] == '_') || (id[0] == '-') || (id[0] == '.')) {
        ++parsed;
        ++id;
    }

    return parsed;
}

/**
 * @brief Parse a node-identifier.
 *
 * node-identifier     = [prefix ":"] identifier
 *
 * @param[in] id Identifier in question.
 * @param[out] prefix Points to the prefix, NULL if there is not any.
 * @param[out] pref_len Length of the prefix, 0 if there is not any.
 * @param[out] name Points to the node name.
 * @param[out] nam_len Length of the node name.
 *
 * @return Number of characters successfully parsed,
 *         positive on success, negative on failure.
 */
int
parse_node_identifier(const char *id, const char **prefix, int *pref_len, const char **name, int *nam_len)
{
    int parsed = 0, ret;

    assert(id);
    if (prefix) {
        *prefix = NULL;
    }
    if (pref_len) {
        *pref_len = 0;
    }
    if (name) {
        *name = NULL;
    }
    if (nam_len) {
        *nam_len = 0;
    }

    if ((ret = parse_identifier(id)) < 1) {
        return ret;
    }

    if (prefix) {
        *prefix = id;
    }
    if (pref_len) {
        *pref_len = ret;
    }

    parsed += ret;
    id += ret;

    /* there is prefix */
    if (id[0] == ':') {
        ++parsed;
        ++id;

    /* there isn't */
    } else {
        if (name && prefix) {
            *name = *prefix;
        }
        if (prefix) {
            *prefix = NULL;
        }

        if (nam_len && pref_len) {
            *nam_len = *pref_len;
        }
        if (pref_len) {
            *pref_len = 0;
        }

        return parsed;
    }

    /* identifier (node name) */
    if ((ret = parse_identifier(id)) < 1) {
        return -parsed+ret;
    }

    if (name) {
        *name = id;
    }
    if (nam_len) {
        *nam_len = ret;
    }

    return parsed+ret;
}

/**
 * @brief Parse a path-predicate (leafref).
 *
 * path-predicate      = "[" *WSP path-equality-expr *WSP "]"
 * path-equality-expr  = node-identifier *WSP "=" *WSP path-key-expr
 *
 * @param[in] id Identifier in question.
 * @param[out] prefix Points to the prefix, NULL if there is not any.
 * @param[out] pref_len Length of the prefix, 0 if there is not any.
 * @param[out] name Points to the node name.
 * @param[out] nam_len Length of the node name.
 * @param[out] path_key_expr Points to the path-key-expr.
 * @param[out] pke_len Length of the path-key-expr.
 * @param[out] has_predicate Flag to mark whether there is another predicate following.
 *
 * @return Number of characters successfully parsed,
 *         positive on success, negative on failure.
 */
int
parse_path_predicate(const char *id, const char **prefix, int *pref_len, const char **name, int *nam_len, const char **path_key_expr, int *pke_len, int *has_predicate)
{
    const char *ptr;
    int parsed = 0, ret;

    assert(id);
    if (prefix) {
        *prefix = NULL;
    }
    if (pref_len) {
        *pref_len = 0;
    }
    if (name) {
        *name = NULL;
    }
    if (nam_len) {
        *nam_len = 0;
    }
    if (path_key_expr) {
        *path_key_expr = NULL;
    }
    if (pke_len) {
        *pke_len = 0;
    }
    if (has_predicate) {
        *has_predicate = 0;
    }

    if (id[0] != '[') {
        return -parsed;
    }

    ++parsed;
    ++id;

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    if ((ret = parse_node_identifier(id, prefix, pref_len, name, nam_len)) < 1) {
        return -parsed+ret;
    }

    parsed += ret;
    id += ret;

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    if (id[0] != '=') {
        return -parsed;
    }

    ++parsed;
    ++id;

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    if ((ptr = strchr(id, ']')) == NULL) {
        return -parsed;
    }

    --ptr;
    while (isspace(ptr[0])) {
        --ptr;
    }
    ++ptr;

    ret = ptr-id;
    if (path_key_expr) {
        *path_key_expr = id;
    }
    if (pke_len) {
        *pke_len = ret;
    }

    parsed += ret;
    id += ret;

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    assert(id[0] == ']');

    if (id[1] == '[') {
        *has_predicate = 1;
    }

    return parsed+1;
}

/**
 * @brief Parse a path-key-expr (leafref). First call parses "current()", all
 *        the ".." and the first node-identifier, other calls parse a single
 *        node-identifier each.
 *
 * path-key-expr       = current-function-invocation *WSP "/" *WSP
 *                       rel-path-keyexpr
 * rel-path-keyexpr    = 1*(".." *WSP "/" *WSP)
 *                       *(node-identifier *WSP "/" *WSP)
 *                       node-identifier
 *
 * @param[in] id Identifier in question.
 * @param[out] prefix Points to the prefix, NULL if there is not any.
 * @param[out] pref_len Length of the prefix, 0 if there is not any.
 * @param[out] name Points to the node name.
 * @param[out] nam_len Length of the node name.
 * @param[out] parent_times Number of ".." in the path. Must be 0 on the first call,
 *                          must not be changed between consecutive calls.
 * @return Number of characters successfully parsed,
 *         positive on success, negative on failure.
 */
int
parse_path_key_expr(const char *id, const char **prefix, int *pref_len, const char **name, int *nam_len, int *parent_times)
{
    int parsed = 0, ret, par_times = 0;

    assert(id);
    assert(parent_times);
    if (prefix) {
        *prefix = NULL;
    }
    if (pref_len) {
        *pref_len = 0;
    }
    if (name) {
        *name = NULL;
    }
    if (nam_len) {
        *nam_len = 0;
    }

    if (!*parent_times) {
        /* current-function-invocation *WSP "/" *WSP rel-path-keyexpr */
        if (strncmp(id, "current()", 9)) {
            return -parsed;
        }

        parsed += 9;
        id += 9;

        while (isspace(id[0])) {
            ++parsed;
            ++id;
        }

        if (id[0] != '/') {
            return -parsed;
        }

        ++parsed;
        ++id;

        while (isspace(id[0])) {
            ++parsed;
            ++id;
        }

        /* rel-path-keyexpr */
        if (strncmp(id, "..", 2)) {
            return -parsed;
        }
        ++par_times;

        parsed += 2;
        id += 2;

        while (isspace(id[0])) {
            ++parsed;
            ++id;
        }
    }

    /* 1*(".." *WSP "/" *WSP) *(node-identifier *WSP "/" *WSP) node-identifier
     *
     * first parent reference with whitespaces already parsed
     */
    if (id[0] != '/') {
        return -parsed;
    }

    ++parsed;
    ++id;

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    while (!strncmp(id, "..", 2) && !*parent_times) {
        ++par_times;

        parsed += 2;
        id += 2;

        while (isspace(id[0])) {
            ++parsed;
            ++id;
        }

        if (id[0] != '/') {
            return -parsed;
        }

        ++parsed;
        ++id;

        while (isspace(id[0])) {
            ++parsed;
            ++id;
        }
    }

    if (!*parent_times) {
        *parent_times = par_times;
    }

    /* all parent references must be parsed at this point */
    if ((ret = parse_node_identifier(id, prefix, pref_len, name, nam_len)) < 1) {
        return -parsed+ret;
    }

    parsed += ret;
    id += ret;

    return parsed;
}

/**
 * @brief Parse path-arg (leafref).
 *
 * path-arg            = absolute-path / relative-path
 * absolute-path       = 1*("/" (node-identifier *path-predicate))
 * relative-path       = 1*(".." "/") descendant-path
 *
 * @param[in] id Identifier in question.
 * @param[out] prefix Points to the prefix, NULL if there is not any.
 * @param[out] pref_len Length of the prefix, 0 if there is not any.
 * @param[out] name Points to the node name.
 * @param[out] nam_len Length of the node name.
 * @param[out] parent_times Number of ".." in the path. Must be 0 on the first call,
 *                          must not be changed between consecutive calls. -1 if the
 *                          path is relative.
 * @param[out] has_predicate Flag to mark whether there is a predicate specified.
 *
 * @return Number of characters successfully parsed,
 *         positive on success, negative on failure.
 */
int
parse_path_arg(const char *id, const char **prefix, int *pref_len, const char **name, int *nam_len, int *parent_times, int *has_predicate)
{
    int parsed = 0, ret, par_times = 0;

    assert(id);
    assert(parent_times);
    if (prefix) {
        *prefix = NULL;
    }
    if (pref_len) {
        *pref_len = 0;
    }
    if (name) {
        *name = NULL;
    }
    if (nam_len) {
        *nam_len = 0;
    }
    if (has_predicate) {
        *has_predicate = 0;
    }

    if (!*parent_times && !strncmp(id, "..", 2)) {
        ++par_times;

        parsed += 2;
        id += 2;

        while (!strncmp(id, "/..", 3)) {
            ++par_times;

            parsed += 3;
            id += 3;
        }
    }

    if (!*parent_times) {
        if (par_times) {
            *parent_times = par_times;
        } else {
            *parent_times = -1;
        }
    }

    if (id[0] != '/') {
        return -parsed;
    }

    /* skip '/' */
    ++parsed;
    ++id;

    /* node-identifier ([prefix:]identifier) */
    if ((ret = parse_node_identifier(id, prefix, pref_len, name, nam_len)) < 1) {
        return -parsed-ret;
    }

    parsed += ret;
    id += ret;

    /* there is no predicate */
    if ((id[0] == '/') || !id[0]) {
        return parsed;
    } else if (id[0] != '[') {
        return -parsed;
    }

    if (has_predicate) {
        *has_predicate = 1;
    }

    return parsed;
}

/**
 * @brief Parse instance-identifier (instance-identifier).
 *
 * instance-identifier = 1*("/" (node-identifier *predicate))
 *
 * @param[in] id Identifier in question.
 * @param[out] prefix Points to the prefix, NULL if there is not any.
 * @param[out] pref_len Length of the prefix, 0 if there is not any.
 * @param[out] name Points to the node name.
 * @param[out] nam_len Length of the node name.
 * @param[out] has_predicate Flag to mark whether there is a predicate specified.
 *
 * @return Number of characters successfully parsed,
 *         positive on success, negative on failure.
 */
int
parse_instance_identifier(const char *id, const char **prefix, int *pref_len, const char **name, int *nam_len, int *has_predicate)
{
    int parsed = 0, ret;

    assert(id);
    if (prefix) {
        *prefix = NULL;
    }
    if (pref_len) {
        *pref_len = 0;
    }
    if (name) {
        *name = NULL;
    }
    if (nam_len) {
        *nam_len = 0;
    }
    if (has_predicate) {
        *has_predicate = 0;
    }

    if (id[0] != '/') {
        return -parsed;
    }

    ++parsed;
    ++id;

    if ((ret = parse_node_identifier(id, prefix, pref_len, name, nam_len)) < 1) {
        return -parsed+ret;
    }

    parsed += ret;
    id += ret;

    if ((id[0] == '[') && has_predicate) {
        *has_predicate = 1;
    }

    return parsed;
}

/**
 * @brief Parse predicate (instance-identifier).
 *
 * predicate           = "[" *WSP (predicate-expr / pos) *WSP "]"
 * predicate-expr      = (node-identifier / ".") *WSP "=" *WSP
 *                       ((DQUOTE string DQUOTE) /
 *                        (SQUOTE string SQUOTE))
 * pos                 = non-negative-integer-value
 *
 * @param[in] id Identifier in question.
 * @param[out] prefix Points to the prefix, NULL if there is not any.
 * @param[out] pref_len Length of the prefix, 0 if there is not any.
 * @param[out] name Points to the node name. Can be identifier (from node-identifier), "." or pos.
 * @param[out] nam_len Length of the node name.
 * @param[out] value Value the node-identifier must have (string from the grammar),
 *                   NULL if there is not any.
 * @param[out] val_len Length of the value, 0 if there is not any.
 * @param[out] has_predicate Flag to mark whether there is a predicate specified.
 *
 * @return Number of characters successfully parsed,
 *         positive on success, negative on failure.
 */
int
parse_predicate(const char *id, const char **prefix, int *pref_len, const char **name, int *nam_len, const char **value, int *val_len, int *has_predicate)
{
    const char *ptr;
    int parsed = 0, ret;
    char quote;

    assert(id);
    if (prefix) {
        *prefix = NULL;
    }
    if (pref_len) {
        *pref_len = 0;
    }
    if (name) {
        *name = NULL;
    }
    if (nam_len) {
        *nam_len = 0;
    }
    if (value) {
        *value = NULL;
    }
    if (val_len) {
        *val_len = 0;
    }
    if (has_predicate) {
        *has_predicate = 0;
    }

    if (id[0] != '[') {
        return -parsed;
    }

    ++parsed;
    ++id;

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    /* pos */
    if (isdigit(id[0])) {
        if (name) {
            *name = id;
        }

        if (id[0] == '0') {
            ++parsed;
            ++id;

            if (isdigit(id[0])) {
                return -parsed;
            }
        }

        while (isdigit(id[0])) {
            ++parsed;
            ++id;
        }

        if (nam_len) {
            *nam_len = id-(*name);
        }

    /* "." */
    } else if (id[0] == '.') {
        if (name) {
            *name = id;
        }
        if (nam_len) {
            *nam_len = 1;
        }

        ++parsed;
        ++id;

    /* node-identifier */
    } else {
        if ((ret = parse_node_identifier(id, prefix, pref_len, name, nam_len)) < 1) {
            return -parsed+ret;
        }
    }

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    if (id[0] != '=') {
        return -parsed;
    }

    ++parsed;
    ++id;

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    /* ((DQUOTE string DQUOTE) / (SQUOTE string SQUOTE)) */
    if ((id[0] == '\"') || (id[0] == '\'')) {
        quote = id[0];

        ++parsed;
        ++id;

        if ((ptr = strchr(id, quote)) == NULL) {
            return -parsed;
        }
        ret = ptr-id;

        if (value) {
            *value = id;
        }
        if (val_len) {
            *val_len = ret;
        }

        parsed += ret+1;
        id += ret+1;
    } else {
        return -parsed;
    }

    while (isspace(id[0])) {
        ++parsed;
        ++id;
    }

    if (id[0] != ']') {
        return -parsed;
    }

    ++parsed;
    ++id;

    if ((id[0] == '[') && has_predicate) {
        *has_predicate = 1;
    }

    return parsed;
}

/**
 * @brief Parse schema-nodeid.
 *
 * schema-nodeid       = absolute-schema-nodeid /
 *                       descendant-schema-nodeid
 * absolute-schema-nodeid = 1*("/" node-identifier)
 * descendant-schema-nodeid =
 *                       node-identifier
 *                       absolute-schema-nodeid
 *
 * @param[in] id Identifier in question.
 * @param[out] prefix Points to the prefix, NULL if there is not any.
 * @param[out] pref_len Length of the prefix, 0 if there is not any.
 * @param[out] name Points to the node name. Can be identifier (from node-identifier), "." or pos.
 * @param[out] nam_len Length of the node name.
 * @param[out] is_relative Flag to mark whether the nodeid is absolute or descendant. Must be -1
 *                         on the first call, must not be changed between consecutive calls.
 *
 * @return Number of characters successfully parsed,
 *         positive on success, negative on failure.
 */
int
parse_schema_nodeid(const char *id, const char **prefix, int *pref_len, const char **name, int *nam_len, int *is_relative)
{
    int parsed = 0, ret;

    assert(id);
    assert(is_relative);
    if (prefix) {
        *prefix = NULL;
    }
    if (pref_len) {
        *pref_len = 0;
    }
    if (name) {
        *name = NULL;
    }
    if (nam_len) {
        *nam_len = 0;
    }

    if (id[0] != '/') {
        if (*is_relative != -1) {
            return -parsed;
        } else {
            *is_relative = 1;
        }
    } else {
        if (*is_relative == -1) {
            *is_relative = 0;
        }
        ++parsed;
        ++id;
    }

    if ((ret = parse_node_identifier(id, prefix, pref_len, name, nam_len)) < 1) {
        return -parsed+ret;
    }

    return parsed+ret;
}