Path of Exile 2 Build Optimizer MCP

""" poe.ninja API Client with Web Scraping Fallback Fetches character data, build rankings, and economy data from poe.ninja """ import httpx import json import logging from typing import Dict, List, Optional, Any from bs4 import BeautifulSoup from datetime import datetime try: from ..api.rate_limiter import RateLimiter from ..api.cache_manager import CacheManager except ImportError: from src.api.rate_limiter import RateLimiter from src.api.cache_manager import CacheManager logger = logging.getLogger(__name__) class PoeNinjaAPI: """ poe.ninja API client with web scraping fallback Fetches character builds, item prices, and meta information """ # League name to URL slug mapping LEAGUE_MAPPINGS = { # Abyss League variants "Rise of the Abyssal": "abyss", "Abyss": "abyss", "Abyss Hardcore": "abysshc", "Abyss HC": "abysshc", "Abyss SSF": "abyssssf", "Abyss HC SSF": "abysshcssf", "Abyss Hardcore SSF": "abysshcssf", # Standard leagues "Standard": "standard", "Hardcore": "hardcore", "SSF Standard": "ssf-standard", "SSF Hardcore": "ssf-hardcore", # Race events (add as discovered) "Act 4 Boss Kill Race 3 SSF": "act4bosskillrace3ssf", } def __init__( self, rate_limiter: Optional[RateLimiter] = None, cache_manager: Optional[CacheManager] = None ): self.base_url = "https://poe.ninja" self.api_base = f"{self.base_url}/api/data" self.rate_limiter = rate_limiter or RateLimiter(rate_limit=20) self.cache_manager = cache_manager self.client = httpx.AsyncClient( timeout=30.0, follow_redirects=True, headers={ "User-Agent": "PoE2-MCP-Server/1.0", "Accept": "application/json, text/html", } ) def _get_league_slug(self, league: str) -> str: """ Convert league name to poe.ninja URL slug Args: league: Full league name (e.g., "Rise of the Abyssal") Returns: URL slug (e.g., "abyss") """ # Check exact match first if league in self.LEAGUE_MAPPINGS: return self.LEAGUE_MAPPINGS[league] # Check case-insensitive match for key, value in self.LEAGUE_MAPPINGS.items(): if key.lower() == league.lower(): return value # Default: convert to lowercase and replace spaces with hyphens return league.lower().replace(" ", "-") async def get_character(self, account: str, character: str, league: str = "Abyss") -> Optional[Dict[str, Any]]: """ Fetch character from poe.ninja using their hidden API Args: account: Path of Exile account name character: Character name league: League name (default: "Abyss") Returns: Character data dictionary or None if not found """ cache_key = f"ninja_character_{account}_{character}_{league}" # Check cache first if self.cache_manager: cached = await self.cache_manager.get(cache_key) if cached: logger.info(f"✅ Cache hit for character {character} ({league})") return cached try: # Rate limit await self.rate_limiter.acquire() logger.info(f"🔍 Fetching character: {character} (Account: {account}, League: {league})") # Use the discovered hidden API endpoint char_data = await self._fetch_character_from_api(account, character, league) if char_data and self.cache_manager: await self.cache_manager.set(cache_key, char_data, ttl=3600) logger.info(f"✅ Successfully fetched and cached character {character}") return char_data except Exception as e: logger.error(f"❌ Error fetching character from poe.ninja: {e}", exc_info=True) return None async def _get_index_state(self) -> Optional[Dict[str, Any]]: """ Fetch the index state which contains snapshot versions for all leagues Returns: Index state with snapshot versions or None if failed """ try: url = f"{self.base_url}/poe2/api/data/index-state" logger.debug(f"Fetching index state from: {url}") response = await self.client.get(url) if response.status_code == 200: data = response.json() logger.debug(f"✅ Got index state with {len(data.get('snapshotVersions', []))} snapshot versions") return data else: logger.warning(f"⚠️ Index state returned {response.status_code}") return None except Exception as e: logger.error(f"❌ Failed to fetch index state: {e}") return None async def _fetch_character_from_api(self, account: str, character: str, league: str) -> Optional[Dict[str, Any]]: """ Fetch character using the discovered hidden API API Endpoint: /poe2/api/builds/{version}/character Parameters: account, name, overview Args: account: Account name character: Character name league: League name Returns: Character data dictionary or None if not found """ try: # Step 1: Get index state to find the snapshot version for this league index_state = await self._get_index_state() if not index_state: logger.warning("⚠️ Could not get index state, falling back to HTML scraping") return await self._scrape_character_page(account, character, league) # Step 2: Find the snapshot version for our league league_slug = self._get_league_slug(league) snapshot = None for snap in index_state.get("snapshotVersions", []): if snap.get("url") == league_slug: snapshot = snap break if not snapshot: logger.warning(f"⚠️ No snapshot found for league '{league}' (slug: '{league_slug}')") logger.warning(f" Available leagues: {[s.get('url') for s in index_state.get('snapshotVersions', [])]}") return await self._scrape_character_page(account, character, league) version = snapshot.get("version") overview = snapshot.get("snapshotName") logger.info(f"📡 Using snapshot version: {version}, overview: {overview}") # Step 3: Call the character API url = f"{self.base_url}/poe2/api/builds/{version}/character" params = { "account": account, "name": character, "overview": overview } logger.debug(f"Calling API: {url}") logger.debug(f"Parameters: {params}") response = await self.client.get(url, params=params) if response.status_code == 200: data = response.json() logger.info(f"✅ Successfully fetched character from API") logger.debug(f" Character: {data.get('name')}, Level: {data.get('level', 'Unknown')}, Class: {data.get('class', 'Unknown')}") return self._normalize_api_character_data(data) elif response.status_code == 404: logger.warning(f"⚠️ Character not found (404)") return None else: logger.warning(f"⚠️ API returned {response.status_code}") logger.debug(f" Response: {response.text[:200]}") return await self._scrape_character_page(account, character, league) except Exception as e: logger.error(f"❌ API fetch failed: {e}", exc_info=True) logger.info(" Falling back to HTML scraping") return await self._scrape_character_page(account, character, league) def _normalize_api_character_data(self, api_data: Dict[str, Any]) -> Dict[str, Any]: """ Normalize character data from the API to match our expected format Args: api_data: Raw data from poe.ninja API Returns: Normalized character data with all stats """ defensive_stats = api_data.get("defensiveStats", {}) # Log what we're receiving for debugging logger.debug(f"🔍 Normalizing API data - defensiveStats: {len(defensive_stats)} fields") logger.debug(f" Life: {defensive_stats.get('life')}, ES: {defensive_stats.get('energyShield')}, EHP: {defensive_stats.get('effectiveHealthPool')}") logger.debug(f" Resistances - Fire: {defensive_stats.get('fireResistance')}, Cold: {defensive_stats.get('coldResistance')}, Lightning: {defensive_stats.get('lightningResistance')}") # Map all defensive stats stats_dict = { # Core defenses "life": defensive_stats.get("life", 0), "energy_shield": defensive_stats.get("energyShield", 0), "mana": defensive_stats.get("mana", 0), "spirit": defensive_stats.get("spirit", 0), "evasion": defensive_stats.get("evasionRating", 0), "armor": defensive_stats.get("armour", 0), # Attributes "strength": defensive_stats.get("strength", 0), "dexterity": defensive_stats.get("dexterity", 0), "intelligence": defensive_stats.get("intelligence", 0), # Resistances "fire_res": defensive_stats.get("fireResistance", 0), "cold_res": defensive_stats.get("coldResistance", 0), "lightning_res": defensive_stats.get("lightningResistance", 0), "chaos_res": defensive_stats.get("chaosResistance", 0), "fire_res_overcap": defensive_stats.get("fireResistanceOverCap", 0), "cold_res_overcap": defensive_stats.get("coldResistanceOverCap", 0), "lightning_res_overcap": defensive_stats.get("lightningResistanceOverCap", 0), "chaos_res_overcap": defensive_stats.get("chaosResistanceOverCap", 0), # EHP and Maximum Hit Taken "effective_health_pool": defensive_stats.get("effectiveHealthPool", 0), "physical_max_hit": defensive_stats.get("physicalMaximumHitTaken", 0), "fire_max_hit": defensive_stats.get("fireMaximumHitTaken", 0), "cold_max_hit": defensive_stats.get("coldMaximumHitTaken", 0), "lightning_max_hit": defensive_stats.get("lightningMaximumHitTaken", 0), "chaos_max_hit": defensive_stats.get("chaosMaximumHitTaken", 0), "lowest_max_hit": defensive_stats.get("lowestMaximumHitTaken", 0), # Charges "endurance_charges": defensive_stats.get("enduranceCharges", 0), "frenzy_charges": defensive_stats.get("frenzyCharges", 0), "power_charges": defensive_stats.get("powerCharges", 0), # Avoidance & Mitigation "block_chance": defensive_stats.get("blockChance", 0), "spell_block_chance": defensive_stats.get("spellBlockChance", 0), "spell_suppression": defensive_stats.get("spellSuppressionChance", 0), "spell_dodge": defensive_stats.get("spellDodgeChance", 0), # Other stats "movement_speed": defensive_stats.get("movementSpeed", 0), "item_rarity": defensive_stats.get("itemRarity", 0), # Physical damage conversion "physical_taken_as": defensive_stats.get("physicalTakenAs", { "physical": 100, "fire": 0, "cold": 0, "lightning": 0, "chaos": 0 }), } # Extract skill DPS data skill_dps = [] for skill in api_data.get("skills", []): for dps_entry in skill.get("dps", []): damage_types = dps_entry.get("damageTypes", [0, 0, 0, 0, 0]) skill_dps.append({ "skill_name": dps_entry.get("name", "Unknown"), "total_dps": dps_entry.get("dps", 0), "dot_dps": dps_entry.get("dotDps", 0), "damage_types": damage_types, "damage_breakdown": { "physical": damage_types[0] if len(damage_types) > 0 else 0, "fire": damage_types[1] if len(damage_types) > 1 else 0, "cold": damage_types[2] if len(damage_types) > 2 else 0, "lightning": damage_types[3] if len(damage_types) > 3 else 0, "chaos": damage_types[4] if len(damage_types) > 4 else 0, } }) # Build normalized data structure normalized = { "name": api_data.get("name", "Unknown"), "account": api_data.get("account", "Unknown"), "class": api_data.get("class", "Unknown"), "level": api_data.get("level", 0), "league": api_data.get("league", "Unknown"), # Items with details "items": [ { "slot": item.get("itemSlot", 0), "name": item.get("itemData", {}).get("name", ""), "type_line": item.get("itemData", {}).get("typeLine", ""), "base_type": item.get("itemData", {}).get("baseType", ""), "item_level": item.get("itemData", {}).get("ilvl", 0), "rarity": item.get("itemData", {}).get("frameType", 0), "corrupted": item.get("itemData", {}).get("corrupted", False), "icon": item.get("itemData", {}).get("icon", ""), "mods": { "implicit": item.get("itemData", {}).get("implicitMods", []), "explicit": item.get("itemData", {}).get("explicitMods", []), "crafted": item.get("itemData", {}).get("craftedMods", []), "enchant": item.get("itemData", {}).get("enchantMods", []) } } for item in api_data.get("items", []) ], # Skills (raw) "skills": api_data.get("skills", []), # Skill DPS (normalized) "skill_dps": skill_dps, # Passives "passives": api_data.get("passiveSelection", []), "passive_set_1": api_data.get("passiveSelectionSet1", []), "passive_set_2": api_data.get("passiveSelectionSet2", []), # Keystones "keystones": [ { "name": keystone.get("name", "Unknown"), "icon": keystone.get("icon", ""), "stats": keystone.get("stats", []) } for keystone in api_data.get("keystones", []) ], # Flasks "flasks": [ { "name": flask.get("itemData", {}).get("typeLine", "Unknown Flask"), "base_type": flask.get("itemData", {}).get("baseType", ""), "item_level": flask.get("itemData", {}).get("ilvl", 0), "mods": flask.get("itemData", {}).get("explicitMods", []), "icon": flask.get("itemData", {}).get("icon", "") } for flask in api_data.get("flasks", []) ], # Jewels "jewels": [ { "name": jewel.get("itemData", {}).get("name") or jewel.get("itemData", {}).get("typeLine", "Unknown Jewel"), "base_type": jewel.get("itemData", {}).get("baseType", ""), "item_level": jewel.get("itemData", {}).get("ilvl", 0), "mods": jewel.get("itemData", {}).get("explicitMods", []), "icon": jewel.get("itemData", {}).get("icon", ""), "position": { "x": jewel.get("itemData", {}).get("x", 0), "y": jewel.get("itemData", {}).get("y", 0) } } for jewel in api_data.get("jewels", []) ], # Path of Building export "pob_export": api_data.get("pathOfBuildingExport", ""), # Stats (nested format) "stats": stats_dict, # Metadata "source": "poe.ninja API", "fetched_at": datetime.utcnow().isoformat(), "weapon_swap_active": api_data.get("useSecondWeaponSet", False) } # ALSO add stats at top level for tools that expect them there normalized.update(stats_dict) logger.info(f"✅ Normalized character data:") logger.info(f" Defenses - Life: {stats_dict.get('life')}, ES: {stats_dict.get('energy_shield')}, EHP: {stats_dict.get('effective_health_pool')}") logger.info(f" Skills with DPS: {len(skill_dps)}") logger.info(f" Keystones: {len(normalized['keystones'])}") logger.info(f" Items: {len(normalized['items'])}, Flasks: {len(normalized['flasks'])}, Jewels: {len(normalized['jewels'])}") return normalized async def _scrape_character_page(self, account: str, character: str, league: str = "Abyss") -> Optional[Dict[str, Any]]: """ Scrape character data from poe.ninja profile page Args: account: Account name character: Character name league: League name (default: "Abyss") Returns: Parsed character data """ try: # Convert league to URL slug (e.g., "Abyss" -> "abyss") league_slug = self._get_league_slug(league) # CRITICAL FIX: Based on HAR file analysis, the correct URL format includes league # Format: https://poe.ninja/poe2/builds/{league}/character/{account}/{character} urls = [ f"{self.base_url}/poe2/builds/{league_slug}/character/{account}/{character}", f"{self.base_url}/builds/{league_slug}/character/{account}/{character}", # Fallback without poe2 ] logger.info(f"📡 Attempting to fetch from poe.ninja with league '{league}' (slug: '{league_slug}')") for i, url in enumerate(urls, 1): try: logger.debug(f" [{i}/{len(urls)}] Trying URL: {url}") response = await self.client.get(url) logger.debug(f" [{i}/{len(urls)}] Response: {response.status_code}") if response.status_code == 200: logger.info(f"✅ Successfully fetched from: {url}") return await self._parse_character_html(response.text, account, character) else: logger.debug(f" [{i}/{len(urls)}] Non-200 status: {response.status_code}") except Exception as e: logger.debug(f" [{i}/{len(urls)}] Exception: {e}") continue logger.warning(f"❌ Could not fetch character {character} from any poe.ninja URL") logger.warning(f" Tried {len(urls)} URLs with league slug '{league_slug}'") return None except Exception as e: logger.error(f"❌ Character scraping error: {e}", exc_info=True) return None async def _parse_character_html(self, html: str, account: str, character: str) -> Optional[Dict[str, Any]]: """ Parse character data from HTML page Args: html: HTML content account: Account name character: Character name Returns: Parsed character data """ try: logger.debug(f"📄 Parsing HTML (length: {len(html)} chars)") soup = BeautifulSoup(html, 'html.parser') # Look for embedded JSON data in script tags scripts = soup.find_all('script') logger.debug(f"🔎 Found {len(scripts)} script tags in HTML") for i, script in enumerate(scripts, 1): if script.string and ('window.__NUXT__' in script.string or 'window.__data' in script.string): try: # Extract JSON data script_content = script.string # Try NUXT data first if 'window.__NUXT__' in script_content: logger.debug(f" [Script {i}] Found window.__NUXT__ data") json_start = script_content.find('window.__NUXT__=') + len('window.__NUXT__=') json_end = script_content.find('</script>', json_start) json_str = script_content[json_start:json_end].strip() if json_str.endswith(';'): json_str = json_str[:-1] logger.debug(f" [Script {i}] Parsing JSON (length: {len(json_str)} chars)") data = json.loads(json_str) logger.info(f"✅ Successfully parsed window.__NUXT__ JSON") return self._extract_character_from_nuxt(data, account, character) # Try __data format elif 'window.__data' in script_content: logger.debug(f" [Script {i}] Found window.__data") json_start = script_content.find('window.__data=') + len('window.__data=') json_end = script_content.find(';', json_start) json_str = script_content[json_start:json_end].strip() logger.debug(f" [Script {i}] Parsing JSON (length: {len(json_str)} chars)") data = json.loads(json_str) logger.info(f"✅ Successfully parsed window.__data JSON") return self._extract_character_from_data(data, account, character) except json.JSONDecodeError as e: logger.warning(f" [Script {i}] Failed to parse JSON: {e}") continue # Fallback: parse HTML structure directly logger.warning(f"⚠️ No embedded JSON found, falling back to HTML parsing") return self._parse_character_from_html(soup, account, character) except Exception as e: logger.error(f"❌ HTML parsing error: {e}", exc_info=True) return None def _extract_character_from_nuxt(self, data: Dict, account: str, character: str) -> Dict[str, Any]: """Extract character data from NUXT format""" try: # Navigate NUXT data structure if 'data' in data: char_data = data['data'][0] if isinstance(data['data'], list) else data['data'] else: char_data = data return { "name": character, "account": account, "class": char_data.get("class", "Unknown"), "level": char_data.get("level", 0), "league": char_data.get("league", "Unknown"), "items": char_data.get("items", []), "skills": char_data.get("skills", []), "passives": char_data.get("passiveSkills", []), "stats": char_data.get("stats", {}), "source": "poe.ninja", "fetched_at": datetime.utcnow().isoformat() } except Exception as e: logger.error(f"Error extracting NUXT data: {e}") return self._create_minimal_character(account, character) def _extract_character_from_data(self, data: Dict, account: str, character: str) -> Dict[str, Any]: """Extract character data from __data format""" return { "name": character, "account": account, "class": data.get("class", "Unknown"), "level": data.get("level", 0), "league": data.get("league", "Unknown"), "items": data.get("items", []), "skills": data.get("skills", []), "passives": data.get("passives", []), "stats": data.get("stats", {}), "source": "poe.ninja", "fetched_at": datetime.utcnow().isoformat() } def _parse_character_from_html(self, soup: BeautifulSoup, account: str, character: str) -> Dict[str, Any]: """Parse character data directly from HTML structure (fallback)""" try: # Try to extract basic info from HTML char_data = self._create_minimal_character(account, character) # Look for character level level_elem = soup.find(class_=['level', 'character-level']) if level_elem: try: char_data["level"] = int(level_elem.text.strip()) except ValueError: pass # Look for character class class_elem = soup.find(class_=['class', 'character-class']) if class_elem: char_data["class"] = class_elem.text.strip() return char_data except Exception as e: logger.error(f"HTML structure parsing error: {e}") return self._create_minimal_character(account, character) def _create_minimal_character(self, account: str, character: str) -> Dict[str, Any]: """Create minimal character data structure""" return { "name": character, "account": account, "class": "Unknown", "level": 0, "league": "Unknown", "items": [], "skills": [], "passives": [], "stats": {}, "source": "poe.ninja (minimal)", "fetched_at": datetime.utcnow().isoformat() } async def get_top_builds( self, league: str = "Standard", class_name: Optional[str] = None, skill: Optional[str] = None, limit: int = 10 ) -> List[Dict[str, Any]]: """ Get top builds from poe.ninja ladder Args: league: League name (e.g., "Rise of the Abyssal", "Standard") class_name: Filter by character class skill: Filter by main skill limit: Maximum number of builds to return Returns: List of build data dictionaries """ # Get the URL slug for this league league_slug = self._get_league_slug(league) cache_key = f"ninja_top_builds_{league_slug}_{class_name}_{skill}_{limit}" if self.cache_manager: cached = await self.cache_manager.get(cache_key) if cached: return cached try: await self.rate_limiter.acquire() # Use league slug in the URL path url = f"{self.base_url}/poe2/builds/{league_slug}" logger.info(f"Fetching top builds from: {url}") response = await self.client.get(url) if response.status_code == 200: builds = await self._parse_builds_page(response.text, class_name, skill, limit) if builds and self.cache_manager: await self.cache_manager.set(cache_key, builds, ttl=1800) logger.info(f"Found {len(builds)} builds from poe.ninja") return builds else: logger.warning(f"poe.ninja builds page returned {response.status_code} for league '{league_slug}'") return [] except Exception as e: logger.error(f"Error fetching top builds: {e}") return [] async def _parse_builds_page( self, html: str, class_filter: Optional[str], skill_filter: Optional[str], limit: int ) -> List[Dict[str, Any]]: """Parse builds from HTML page (NUXT data extraction)""" try: soup = BeautifulSoup(html, 'html.parser') builds = [] # poe.ninja uses NUXT, so data is embedded in JavaScript # Look for __NUXT__ data for script in soup.find_all('script'): script_content = script.string if not script_content: continue # Try to find NUXT data if 'window.__NUXT__' in script_content or '__NUXT__=' in script_content: try: # Extract JSON from the script start_marker = '__NUXT__=' if start_marker in script_content: json_start = script_content.find(start_marker) + len(start_marker) # Find the end - it's usually a semicolon or end of script json_end = script_content.find('</script>', json_start) if json_end == -1: json_end = len(script_content) json_str = script_content[json_start:json_end].strip() if json_str.endswith(';'): json_str = json_str[:-1] # Parse the NUXT data nuxt_data = json.loads(json_str) builds = self._extract_builds_from_nuxt(nuxt_data, class_filter, skill_filter, limit) if builds: return builds except json.JSONDecodeError as e: logger.debug(f"Failed to parse NUXT data: {e}") continue # Fallback: Try to find build data in alternative locations # Some pages might have data in different formats logger.warning("Could not find NUXT data, trying HTML fallback") # Look for build listings in HTML build_elements = soup.find_all(class_=['build-row', 'build-item', 'character-row']) for elem in build_elements[:limit * 2]: # Get extra in case of filtering build = self._extract_build_info(elem) if build: # Apply filters if class_filter and build.get("class") != class_filter: continue if skill_filter and skill_filter.lower() not in build.get("main_skill", "").lower(): continue builds.append(build) if len(builds) >= limit: break return builds except Exception as e: logger.error(f"Build parsing error: {e}") return [] def _extract_builds_from_nuxt( self, nuxt_data: Dict, class_filter: Optional[str], skill_filter: Optional[str], limit: int ) -> List[Dict[str, Any]]: """Extract build data from NUXT structure""" builds = [] try: # NUXT data structure varies, but typically: # __NUXT__.data[0] or __NUXT__.state # Navigate through the data structure to find builds/characters # Try different paths data_sources = [ nuxt_data.get('data', []), nuxt_data.get('state', {}).get('builds', []), nuxt_data.get('state', {}).get('characters', []), ] # Also check nested structures if isinstance(nuxt_data, dict): for key in nuxt_data: val = nuxt_data[key] if isinstance(val, list) and len(val) > 0: # Check if this looks like build data if isinstance(val[0], dict) and ('character' in val[0] or 'name' in val[0]): data_sources.append(val) for data_source in data_sources: if not data_source: continue # Handle list of builds if isinstance(data_source, list): for item in data_source: if isinstance(item, dict): build = self._normalize_build_data(item) if build: # Apply filters if class_filter and build.get("class", "").lower() != class_filter.lower(): continue if skill_filter and skill_filter.lower() not in build.get("main_skill", "").lower(): continue builds.append(build) if len(builds) >= limit: return builds # Handle nested structure elif isinstance(data_source, dict): for key, value in data_source.items(): if isinstance(value, list): for item in value: if isinstance(item, dict): build = self._normalize_build_data(item) if build: # Apply filters if class_filter and build.get("class", "").lower() != class_filter.lower(): continue if skill_filter and skill_filter.lower() not in build.get("main_skill", "").lower(): continue builds.append(build) if len(builds) >= limit: return builds except Exception as e: logger.error(f"Error extracting builds from NUXT data: {e}") return builds def _normalize_build_data(self, raw_data: Dict) -> Optional[Dict[str, Any]]: """Normalize build data from various sources""" try: # Try to extract common fields build = { "account": raw_data.get("account", raw_data.get("accountName", "")), "character": raw_data.get("character", raw_data.get("name", raw_data.get("characterName", ""))), "class": raw_data.get("class", raw_data.get("className", raw_data.get("ascendancy", ""))), "level": raw_data.get("level", 0), "main_skill": raw_data.get("mainSkill", raw_data.get("skill", "")), "dps": raw_data.get("dps", 0), } # Skip if we don't have at least character name if not build["character"]: return None return build except Exception as e: logger.debug(f"Failed to normalize build data: {e}") return None def _extract_build_info(self, element) -> Optional[Dict[str, Any]]: """Extract build information from HTML element""" try: build = { "account": element.get("data-account", ""), "character": element.get("data-character", ""), "class": "", "level": 0, "main_skill": "", "dps": 0 } # Try to extract from data attributes or text content class_elem = element.find(class_=['class', 'build-class']) if class_elem: build["class"] = class_elem.text.strip() level_elem = element.find(class_=['level', 'build-level']) if level_elem: try: build["level"] = int(level_elem.text.strip()) except ValueError: pass return build if build["account"] or build["character"] else None except Exception as e: logger.debug(f"Failed to extract build info: {e}") return None async def get_item_prices(self, league: str = "Standard", item_type: str = "UniqueWeapon") -> List[Dict[str, Any]]: """ Get item prices from poe.ninja economy API Args: league: League name item_type: Type of items (UniqueWeapon, UniqueArmour, etc.) Returns: List of items with prices """ cache_key = f"ninja_prices_{league}_{item_type}" if self.cache_manager: cached = await self.cache_manager.get(cache_key) if cached: return cached try: await self.rate_limiter.acquire() url = f"{self.api_base}/itemoverview" params = { "league": league, "type": item_type } response = await self.client.get(url, params=params) if response.status_code == 200: data = response.json() items = data.get("lines", []) if items and self.cache_manager: await self.cache_manager.set(cache_key, items, ttl=3600) return items return [] except Exception as e: logger.error(f"Error fetching item prices: {e}") return [] async def get_pob_import(self, account: str, character: str) -> Optional[str]: """ Get Path of Building import code for a character using poe.ninja's hidden API This endpoint returns a PoB code that can be imported into Path of Building Args: account: Path of Exile account name character: Character name Returns: Base64-encoded PoB build code or None if not found Example: >>> api = PoeNinjaAPI() >>> pob_code = await api.get_pob_import("Tomawar40-2671", "DoesFireWorkGoodNow") >>> print(pob_code) 'eJyLjgUAARUAuQ==' # Base64 PoB code """ cache_key = f"ninja_pob_{account}_{character}" # Check cache first if self.cache_manager: cached = await self.cache_manager.get(cache_key) if cached: logger.info(f"✅ Cache hit for PoB code: {character}") return cached try: # Rate limit await self.rate_limiter.acquire() logger.info(f"📦 Fetching PoB code for character: {character} (Account: {account})") # Call the discovered PoB import API url = f"{self.base_url}/poe2/api/builds/pob/import" params = { "accountName": account, "characterName": character } logger.debug(f"Calling PoB API: {url}") logger.debug(f"Parameters: {params}") # Add referer header to appear as if coming from character page headers = { "Referer": f"{self.base_url}/poe2/builds/character/{account}/{character}", "Accept": "application/json", } response = await self.client.get(url, params=params, headers=headers) if response.status_code == 200: data = response.json() # The API should return a PoB code # Based on typical poe.ninja API structure, it might be in data['pob'] or data['code'] pob_code = data.get("pob") or data.get("code") or data.get("build") if pob_code: logger.info(f"✅ Successfully fetched PoB code for {character}") if self.cache_manager: await self.cache_manager.set(cache_key, pob_code, ttl=3600) return pob_code else: logger.warning(f"⚠️ PoB API returned success but no code found") logger.debug(f" Response data keys: {list(data.keys())}") # Return the full data in case it's in a different format return data elif response.status_code == 404: logger.warning(f"⚠️ Character not found for PoB import (404)") return None else: logger.warning(f"⚠️ PoB API returned {response.status_code}") logger.debug(f" Response: {response.text[:200]}") return None except Exception as e: logger.error(f"❌ PoB import API failed: {e}", exc_info=True) return None async def close(self): """Close HTTP client""" await self.client.aclose()