EvoBGP/internal/pipeline/parse.go

package pipeline

import (
	"bufio"
	"encoding/json"
	"fmt"
	"net/netip"
	"strings"
)

// Защита от pathological JSON: глубокая рекурсия при обходе и взрыв числа узлов по пути.
const (
	maxJSONWalkDepth   = 512
	maxJSONPathBreadth = 50000
)

// ParseCIDRLines extracts unique IPv4/IPv6 CIDRs from plain text (one per line, # comments, empty lines skipped).
func ParseCIDRLines(body string) []netip.Prefix {
	seen := make(map[string]struct{})
	var out []netip.Prefix
	sc := bufio.NewScanner(strings.NewReader(body))
	for sc.Scan() {
		line := strings.TrimSpace(sc.Text())
		if line == "" || strings.HasPrefix(line, "#") {
			continue
		}
		pfx := parseOneCIDR(line)
		if !pfx.IsValid() {
			continue
		}
		m := pfx.Masked()
		s := m.String()
		if _, ok := seen[s]; ok {
			continue
		}
		seen[s] = struct{}{}
		out = append(out, m)
	}
	return out
}

// ExtractCIDRs parses CIDRs from either plaintext lines or JSON payload.
// For sourceKind="json", prefixPath supports dotted traversal, with [] for arrays:
// e.g. "prefixes[]", "data.items[].cidr".
func ExtractCIDRs(body, sourceKind, prefixPath string) ([]netip.Prefix, error) {
	if strings.EqualFold(strings.TrimSpace(sourceKind), "json") {
		return parseCIDRsFromJSON(body, prefixPath)
	}
	return ParseCIDRLines(body), nil
}

func parseCIDRsFromJSON(body, prefixPath string) ([]netip.Prefix, error) {
	var root any
	if err := json.Unmarshal([]byte(body), &root); err != nil {
		return nil, err
	}
	values, err := jsonValuesAtPath(root, prefixPath)
	if err != nil {
		return nil, err
	}
	seen := make(map[string]struct{})
	var out []netip.Prefix
	for _, raw := range values {
		pfx := parseOneCIDR(raw)
		if !pfx.IsValid() {
			continue
		}
		m := pfx.Masked()
		s := m.String()
		if _, ok := seen[s]; ok {
			continue
		}
		seen[s] = struct{}{}
		out = append(out, m)
	}
	return out, nil
}

func jsonValuesAtPath(root any, prefixPath string) ([]string, error) {
	path := strings.TrimSpace(prefixPath)
	if path == "" {
		return flattenJSONStrings(root, 0)
	}
	parts := strings.Split(path, ".")
	nodes := []any{root}
	for _, p := range parts {
		part := strings.TrimSpace(p)
		if part == "" {
			continue
		}
		iter := strings.HasSuffix(part, "[]")
		key := strings.TrimSuffix(part, "[]")
		var next []any
		for _, n := range nodes {
			obj, ok := n.(map[string]any)
			if !ok {
				continue
			}
			child, ok := obj[key]
			if !ok {
				continue
			}
			if iter {
				if arr, ok := child.([]any); ok {
					next = append(next, arr...)
				}
				continue
			}
			next = append(next, child)
		}
		if len(next) > maxJSONPathBreadth {
			return nil, fmt.Errorf("json path: слишком много узлов на шаге (>%d)", maxJSONPathBreadth)
		}
		nodes = next
		if len(nodes) == 0 {
			return nil, nil
		}
	}
	var out []string
	for _, n := range nodes {
		part, err := flattenJSONStrings(n, 0)
		if err != nil {
			return nil, err
		}
		out = append(out, part...)
	}
	return out, nil
}

func flattenJSONStrings(v any, depth int) ([]string, error) {
	if depth > maxJSONWalkDepth {
		return nil, fmt.Errorf("json: глубина вложенности превышает %d", maxJSONWalkDepth)
	}
	switch x := v.(type) {
	case string:
		return []string{strings.TrimSpace(x)}, nil
	case []any:
		var out []string
		for _, item := range x {
			part, err := flattenJSONStrings(item, depth+1)
			if err != nil {
				return nil, err
			}
			out = append(out, part...)
		}
		return out, nil
	case map[string]any:
		var out []string
		for _, item := range x {
			part, err := flattenJSONStrings(item, depth+1)
			if err != nil {
				return nil, err
			}
			out = append(out, part...)
		}
		return out, nil
	default:
		return nil, nil
	}
}

func parseOneCIDR(s string) netip.Prefix {
	if p, err := netip.ParsePrefix(s); err == nil {
		return p
	}
	if addr, err := netip.ParseAddr(s); err == nil {
		if addr.Is4() {
			p, _ := addr.Prefix(32)
			return p
		}
		p, _ := addr.Prefix(128)
		return p
	}
	return netip.Prefix{}
}