XML Parser Project

How to run

In terminal navigate to this directory and enter commands:

 $ make

$ ./xml_parser

You can change the value of the mock input variable *xml_data in main.c to test different inputs.

Overview

This project addresses the challenge of parsing large XML documents where performance is critical. The goal is to build a forward-only XML parser that processes XML efficiently, providing access to data as it traverses the document in a single pass.

Core Requirements

1. Forward-Only Parsing: Traverse the XML document once without revisiting nodes.
2. Event-Based Reporting: Report the current path, element attributes, and values in real-time.
3. Efficient Handling:
   • Identify and decode XML escaped entities (e.g., ", &, etc.).
   • Ignore comments.
4. Support Path and Value Tracking:
   • Include the full XML path (e.g., /root/order/amount).
   • Report attributes and text values at the current path.
5. Example Use Case: Output the order id for any order with an amount > 100.

---

Design and Implementation

1. State Machine

The parser is implemented as a state machine that traverses the XML document. Key states include:

• Start Tag: Handles opening tags and parses attributes.
• End Tag: Handles closing tags.
• Text Content: Extracts and decodes text between tags.
• Comments: Ignores XML comments (<!-- ... -->).

2. Path and Attribute Tracking

• Path Maintenance: Tracks the current XML path and updates it dynamically as elements are encountered.
• Attributes: Parses and stores attributes for the current element, making them available during parsing.

3. XML Entity Decoding

Handles standard XML entities:

• " -> "
• & -> &
• ' -> '
• < -> <
• > -> >

4. Event Reporting

The parser triggers callback functions to report events such as:

• start_element: When an element starts.
• end_element: When an element ends.
• text: When text content is encountered.
• attribute: When attributes are parsed.

---

Example Code

XML Document

<root>
    <order id="1111">
        <amount>150</amount>
    </order>
    <order id="222">
        <amount>2</amount>
    </order>
    <order id="333">
        <amount>2000</amount>
    </order>
</root>

Example Usage

Parser Callback

void example_callback(const char *event, const char *path, const char *key, const char *value) {
    static char current_id[128];
    if (strcmp(event, "attribute") == 0 && strcmp(key, "id") == 0) {
        strcpy(current_id, value); // Store the order id
    } else if (strcmp(event, "text") == 0 && strstr(path, "/order/amount") != NULL) {
        int amount = atoi(value);
        if (amount > 100) {
            printf("Order ID: %s\tAmount: %d\n", current_id, amount);
        }
    }
}

Main Function

int main() {
    const char *xml_data =
        "<root>"
        "    <order id=\"1111\">"
        "        <amount>150</amount>"
        "    </order>"
        "    <order id=\"222\">"
        "        <amount>2</amount>"
        "    </order>"
        "    <order id=\"333\">"
        "        <amount>2000</amount>"
        "    </order>"
        "</root>";

    XMLParser parser;
    init_parser(&parser); // Initialize the parser
    parse_xml(&parser, xml_data, example_callback); // Parse the XML with the callback

    return 0;
}

Expected Output

For the example XML provided, the program should output:

Order ID: 111   Amount: 150
Order ID: 222   Amount: 2
Order ID: 333   Amount: 2000
Order ID: 123   Amount: 15
Order ID: 321   Amount: 4
Order ID: 231   Amount: 200
Order ID: 213   Amount: 200

Summary

This project provides a forward-only XML parser that efficiently processes XML documents in a single pass. The parser is lightweight, fast, and ideal for scenarios where only immediate access to current data is required.

By implementing this parser, you can handle large-scale XML parsing tasks with minimal overhead while ensuring real-time reporting of attributes, paths, and text content.