bytekit-sdk 0.3.0__py3-none-any.whl

bytekit/models/sitemap_request_process_type_0_options.py

@@ -0,0 +1,44 @@
+from collections.abc import Mapping
+from typing import Any, TypeVar
+
+from attrs import define as _attrs_define
+from attrs import field as _attrs_field
+
+T = TypeVar("T", bound="SitemapRequestProcessType0Options")
+
+
+@_attrs_define
+class SitemapRequestProcessType0Options:
+    """ """
+
+    additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
+
+    def to_dict(self) -> dict[str, Any]:
+        field_dict: dict[str, Any] = {}
+        field_dict.update(self.additional_properties)
+
+        return field_dict
+
+    @classmethod
+    def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
+        d = dict(src_dict)
+        sitemap_request_process_type_0_options = cls()
+
+        sitemap_request_process_type_0_options.additional_properties = d
+        return sitemap_request_process_type_0_options
+
+    @property
+    def additional_keys(self) -> list[str]:
+        return list(self.additional_properties.keys())
+
+    def __getitem__(self, key: str) -> Any:
+        return self.additional_properties[key]
+
+    def __setitem__(self, key: str, value: Any) -> None:
+        self.additional_properties[key] = value
+
+    def __delitem__(self, key: str) -> None:
+        del self.additional_properties[key]
+
+    def __contains__(self, key: str) -> bool:
+        return key in self.additional_properties

bytekit/models/sitemap_request_strategy.py

@@ -0,0 +1,10 @@
+from enum import Enum
+
+
+class SitemapRequestStrategy(str, Enum):
+    DEEP = "deep"
+    QUICK = "quick"
+    STANDARD = "standard"
+
+    def __str__(self) -> str:
+        return str(self.value)

bytekit/models/sitemap_response.py

@@ -0,0 +1,418 @@
+import datetime
+from collections.abc import Mapping
+from typing import (
+    TYPE_CHECKING,
+    Any,
+    TypeVar,
+    Union,
+    cast,
+)
+
+from attrs import define as _attrs_define
+from attrs import field as _attrs_field
+from dateutil.parser import isoparse
+
+from ..models.sitemap_response_cache import SitemapResponseCache
+from ..models.sitemap_response_status import SitemapResponseStatus
+from ..types import UNSET, Unset
+
+if TYPE_CHECKING:
+    from ..models.sitemap_response_metadata_type_0 import SitemapResponseMetadataType0
+    from ..models.sitemap_response_process_type_0 import SitemapResponseProcessType0
+    from ..models.sitemap_response_sources_type_0 import SitemapResponseSourcesType0
+
+
+T = TypeVar("T", bound="SitemapResponse")
+
+
+@_attrs_define
+class SitemapResponse:
+    """
+    Attributes:
+        id (str):  Example: sm_01j9abc123.
+        url (str):
+        strategy (str):
+        max_depth (int):
+        max_urls (int):
+        status (SitemapResponseStatus):
+        total_urls (Union[None, int]):
+        sources (Union['SitemapResponseSourcesType0', None]):
+        total_bytes (Union[None, int]):
+        results_url (Union[None, str]): Signed S3 URL for downloading the discovered URLs. Present when
+            completed. The payload is a gzip-compressed CSV with a two-column
+            `url,source` header (one row per discovered URL); `source` is one of
+            `sitemap` (found in a sitemap.xml), `known` (seen in prior scrape
+            history), or `links` (discovered by crawling page links). In
+            `compact: true` mode the payload is a plain newline-delimited URL
+            list with no header. `lastmod`/`depth` columns are not emitted.
+        warnings (Union[None, list[str]]):
+        error_code (Union[None, str]):
+        error_message (Union[None, str]):
+        compact (bool):
+        process (Union['SitemapResponseProcessType0', None]):
+        metadata (Union['SitemapResponseMetadataType0', None]):
+        created_at (datetime.datetime):
+        completed_at (Union[None, datetime.datetime]):
+        expires_at (Union[None, datetime.datetime]):
+        cache (Union[Unset, SitemapResponseCache]): Cache verdict for the request that produced this creation response.
+            `hit` — served from a completed cached clone of a prior matching
+            sitemap job (billed at 0.5×). `miss` — no fresh cache entry; a new
+            discovery was queued (the result is not yet cached). `bypass` —
+            `cache_ttl: '0'` skipped the cache read. Present on the 202 creation
+            response; omitted on `GET /v1/sitemap/{id}` and any response with no
+            cache verdict. Optional — pre-cache-era consumers may not observe it.
+        cache_age_s (Union[None, Unset, int]): Age in whole seconds of the cached source job (its `completed_at`)
+            replayed on a `cache: hit`. Clamped to ≥ 0 to absorb clock skew. Present only on `cache: hit`; null on
+            `miss`/`bypass`.
+    """
+
+    id: str
+    url: str
+    strategy: str
+    max_depth: int
+    max_urls: int
+    status: SitemapResponseStatus
+    total_urls: Union[None, int]
+    sources: Union["SitemapResponseSourcesType0", None]
+    total_bytes: Union[None, int]
+    results_url: Union[None, str]
+    warnings: Union[None, list[str]]
+    error_code: Union[None, str]
+    error_message: Union[None, str]
+    compact: bool
+    process: Union["SitemapResponseProcessType0", None]
+    metadata: Union["SitemapResponseMetadataType0", None]
+    created_at: datetime.datetime
+    completed_at: Union[None, datetime.datetime]
+    expires_at: Union[None, datetime.datetime]
+    cache: Union[Unset, SitemapResponseCache] = UNSET
+    cache_age_s: Union[None, Unset, int] = UNSET
+    additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
+
+    def to_dict(self) -> dict[str, Any]:
+        from ..models.sitemap_response_metadata_type_0 import (
+            SitemapResponseMetadataType0,
+        )
+        from ..models.sitemap_response_process_type_0 import SitemapResponseProcessType0
+        from ..models.sitemap_response_sources_type_0 import SitemapResponseSourcesType0
+
+        id = self.id
+
+        url = self.url
+
+        strategy = self.strategy
+
+        max_depth = self.max_depth
+
+        max_urls = self.max_urls
+
+        status = self.status.value
+
+        total_urls: Union[None, int]
+        total_urls = self.total_urls
+
+        sources: Union[None, dict[str, Any]]
+        if isinstance(self.sources, SitemapResponseSourcesType0):
+            sources = self.sources.to_dict()
+        else:
+            sources = self.sources
+
+        total_bytes: Union[None, int]
+        total_bytes = self.total_bytes
+
+        results_url: Union[None, str]
+        results_url = self.results_url
+
+        warnings: Union[None, list[str]]
+        if isinstance(self.warnings, list):
+            warnings = self.warnings
+
+        else:
+            warnings = self.warnings
+
+        error_code: Union[None, str]
+        error_code = self.error_code
+
+        error_message: Union[None, str]
+        error_message = self.error_message
+
+        compact = self.compact
+
+        process: Union[None, dict[str, Any]]
+        if isinstance(self.process, SitemapResponseProcessType0):
+            process = self.process.to_dict()
+        else:
+            process = self.process
+
+        metadata: Union[None, dict[str, Any]]
+        if isinstance(self.metadata, SitemapResponseMetadataType0):
+            metadata = self.metadata.to_dict()
+        else:
+            metadata = self.metadata
+
+        created_at = self.created_at.isoformat()
+
+        completed_at: Union[None, str]
+        if isinstance(self.completed_at, datetime.datetime):
+            completed_at = self.completed_at.isoformat()
+        else:
+            completed_at = self.completed_at
+
+        expires_at: Union[None, str]
+        if isinstance(self.expires_at, datetime.datetime):
+            expires_at = self.expires_at.isoformat()
+        else:
+            expires_at = self.expires_at
+
+        cache: Union[Unset, str] = UNSET
+        if not isinstance(self.cache, Unset):
+            cache = self.cache.value
+
+        cache_age_s: Union[None, Unset, int]
+        if isinstance(self.cache_age_s, Unset):
+            cache_age_s = UNSET
+        else:
+            cache_age_s = self.cache_age_s
+
+        field_dict: dict[str, Any] = {}
+        field_dict.update(self.additional_properties)
+        field_dict.update(
+            {
+                "id": id,
+                "url": url,
+                "strategy": strategy,
+                "max_depth": max_depth,
+                "max_urls": max_urls,
+                "status": status,
+                "total_urls": total_urls,
+                "sources": sources,
+                "total_bytes": total_bytes,
+                "results_url": results_url,
+                "warnings": warnings,
+                "error_code": error_code,
+                "error_message": error_message,
+                "compact": compact,
+                "process": process,
+                "metadata": metadata,
+                "created_at": created_at,
+                "completed_at": completed_at,
+                "expires_at": expires_at,
+            }
+        )
+        if cache is not UNSET:
+            field_dict["cache"] = cache
+        if cache_age_s is not UNSET:
+            field_dict["cache_age_s"] = cache_age_s
+
+        return field_dict
+
+    @classmethod
+    def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
+        from ..models.sitemap_response_metadata_type_0 import (
+            SitemapResponseMetadataType0,
+        )
+        from ..models.sitemap_response_process_type_0 import SitemapResponseProcessType0
+        from ..models.sitemap_response_sources_type_0 import SitemapResponseSourcesType0
+
+        d = dict(src_dict)
+        id = d.pop("id")
+
+        url = d.pop("url")
+
+        strategy = d.pop("strategy")
+
+        max_depth = d.pop("max_depth")
+
+        max_urls = d.pop("max_urls")
+
+        status = SitemapResponseStatus(d.pop("status"))
+
+        def _parse_total_urls(data: object) -> Union[None, int]:
+            if data is None:
+                return data
+            return cast(Union[None, int], data)
+
+        total_urls = _parse_total_urls(d.pop("total_urls"))
+
+        def _parse_sources(data: object) -> Union["SitemapResponseSourcesType0", None]:
+            if data is None:
+                return data
+            try:
+                if not isinstance(data, dict):
+                    raise TypeError()
+                sources_type_0 = SitemapResponseSourcesType0.from_dict(data)
+
+                return sources_type_0
+            except:  # noqa: E722
+                pass
+            return cast(Union["SitemapResponseSourcesType0", None], data)
+
+        sources = _parse_sources(d.pop("sources"))
+
+        def _parse_total_bytes(data: object) -> Union[None, int]:
+            if data is None:
+                return data
+            return cast(Union[None, int], data)
+
+        total_bytes = _parse_total_bytes(d.pop("total_bytes"))
+
+        def _parse_results_url(data: object) -> Union[None, str]:
+            if data is None:
+                return data
+            return cast(Union[None, str], data)
+
+        results_url = _parse_results_url(d.pop("results_url"))
+
+        def _parse_warnings(data: object) -> Union[None, list[str]]:
+            if data is None:
+                return data
+            try:
+                if not isinstance(data, list):
+                    raise TypeError()
+                warnings_type_0 = cast(list[str], data)
+
+                return warnings_type_0
+            except:  # noqa: E722
+                pass
+            return cast(Union[None, list[str]], data)
+
+        warnings = _parse_warnings(d.pop("warnings"))
+
+        def _parse_error_code(data: object) -> Union[None, str]:
+            if data is None:
+                return data
+            return cast(Union[None, str], data)
+
+        error_code = _parse_error_code(d.pop("error_code"))
+
+        def _parse_error_message(data: object) -> Union[None, str]:
+            if data is None:
+                return data
+            return cast(Union[None, str], data)
+
+        error_message = _parse_error_message(d.pop("error_message"))
+
+        compact = d.pop("compact")
+
+        def _parse_process(data: object) -> Union["SitemapResponseProcessType0", None]:
+            if data is None:
+                return data
+            try:
+                if not isinstance(data, dict):
+                    raise TypeError()
+                process_type_0 = SitemapResponseProcessType0.from_dict(data)
+
+                return process_type_0
+            except:  # noqa: E722
+                pass
+            return cast(Union["SitemapResponseProcessType0", None], data)
+
+        process = _parse_process(d.pop("process"))
+
+        def _parse_metadata(
+            data: object,
+        ) -> Union["SitemapResponseMetadataType0", None]:
+            if data is None:
+                return data
+            try:
+                if not isinstance(data, dict):
+                    raise TypeError()
+                metadata_type_0 = SitemapResponseMetadataType0.from_dict(data)
+
+                return metadata_type_0
+            except:  # noqa: E722
+                pass
+            return cast(Union["SitemapResponseMetadataType0", None], data)
+
+        metadata = _parse_metadata(d.pop("metadata"))
+
+        created_at = isoparse(d.pop("created_at"))
+
+        def _parse_completed_at(data: object) -> Union[None, datetime.datetime]:
+            if data is None:
+                return data
+            try:
+                if not isinstance(data, str):
+                    raise TypeError()
+                completed_at_type_0 = isoparse(data)
+
+                return completed_at_type_0
+            except:  # noqa: E722
+                pass
+            return cast(Union[None, datetime.datetime], data)
+
+        completed_at = _parse_completed_at(d.pop("completed_at"))
+
+        def _parse_expires_at(data: object) -> Union[None, datetime.datetime]:
+            if data is None:
+                return data
+            try:
+                if not isinstance(data, str):
+                    raise TypeError()
+                expires_at_type_0 = isoparse(data)
+
+                return expires_at_type_0
+            except:  # noqa: E722
+                pass
+            return cast(Union[None, datetime.datetime], data)
+
+        expires_at = _parse_expires_at(d.pop("expires_at"))
+
+        _cache = d.pop("cache", UNSET)
+        cache: Union[Unset, SitemapResponseCache]
+        if isinstance(_cache, Unset):
+            cache = UNSET
+        else:
+            cache = SitemapResponseCache(_cache)
+
+        def _parse_cache_age_s(data: object) -> Union[None, Unset, int]:
+            if data is None:
+                return data
+            if isinstance(data, Unset):
+                return data
+            return cast(Union[None, Unset, int], data)
+
+        cache_age_s = _parse_cache_age_s(d.pop("cache_age_s", UNSET))
+
+        sitemap_response = cls(
+            id=id,
+            url=url,
+            strategy=strategy,
+            max_depth=max_depth,
+            max_urls=max_urls,
+            status=status,
+            total_urls=total_urls,
+            sources=sources,
+            total_bytes=total_bytes,
+            results_url=results_url,
+            warnings=warnings,
+            error_code=error_code,
+            error_message=error_message,
+            compact=compact,
+            process=process,
+            metadata=metadata,
+            created_at=created_at,
+            completed_at=completed_at,
+            expires_at=expires_at,
+            cache=cache,
+            cache_age_s=cache_age_s,
+        )
+
+        sitemap_response.additional_properties = d
+        return sitemap_response
+
+    @property
+    def additional_keys(self) -> list[str]:
+        return list(self.additional_properties.keys())
+
+    def __getitem__(self, key: str) -> Any:
+        return self.additional_properties[key]
+
+    def __setitem__(self, key: str, value: Any) -> None:
+        self.additional_properties[key] = value
+
+    def __delitem__(self, key: str) -> None:
+        del self.additional_properties[key]
+
+    def __contains__(self, key: str) -> bool:
+        return key in self.additional_properties

bytekit/models/sitemap_response_cache.py

@@ -0,0 +1,10 @@
+from enum import Enum
+
+
+class SitemapResponseCache(str, Enum):
+    BYPASS = "bypass"
+    HIT = "hit"
+    MISS = "miss"
+
+    def __str__(self) -> str:
+        return str(self.value)

bytekit/models/sitemap_response_metadata_type_0.py

@@ -0,0 +1,44 @@
+from collections.abc import Mapping
+from typing import Any, TypeVar
+
+from attrs import define as _attrs_define
+from attrs import field as _attrs_field
+
+T = TypeVar("T", bound="SitemapResponseMetadataType0")
+
+
+@_attrs_define
+class SitemapResponseMetadataType0:
+    """ """
+
+    additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
+
+    def to_dict(self) -> dict[str, Any]:
+        field_dict: dict[str, Any] = {}
+        field_dict.update(self.additional_properties)
+
+        return field_dict
+
+    @classmethod
+    def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
+        d = dict(src_dict)
+        sitemap_response_metadata_type_0 = cls()
+
+        sitemap_response_metadata_type_0.additional_properties = d
+        return sitemap_response_metadata_type_0
+
+    @property
+    def additional_keys(self) -> list[str]:
+        return list(self.additional_properties.keys())
+
+    def __getitem__(self, key: str) -> Any:
+        return self.additional_properties[key]
+
+    def __setitem__(self, key: str, value: Any) -> None:
+        self.additional_properties[key] = value
+
+    def __delitem__(self, key: str) -> None:
+        del self.additional_properties[key]
+
+    def __contains__(self, key: str) -> bool:
+        return key in self.additional_properties

bytekit/models/sitemap_response_process_type_0.py

@@ -0,0 +1,44 @@
+from collections.abc import Mapping
+from typing import Any, TypeVar
+
+from attrs import define as _attrs_define
+from attrs import field as _attrs_field
+
+T = TypeVar("T", bound="SitemapResponseProcessType0")
+
+
+@_attrs_define
+class SitemapResponseProcessType0:
+    """ """
+
+    additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
+
+    def to_dict(self) -> dict[str, Any]:
+        field_dict: dict[str, Any] = {}
+        field_dict.update(self.additional_properties)
+
+        return field_dict
+
+    @classmethod
+    def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
+        d = dict(src_dict)
+        sitemap_response_process_type_0 = cls()
+
+        sitemap_response_process_type_0.additional_properties = d
+        return sitemap_response_process_type_0
+
+    @property
+    def additional_keys(self) -> list[str]:
+        return list(self.additional_properties.keys())
+
+    def __getitem__(self, key: str) -> Any:
+        return self.additional_properties[key]
+
+    def __setitem__(self, key: str, value: Any) -> None:
+        self.additional_properties[key] = value
+
+    def __delitem__(self, key: str) -> None:
+        del self.additional_properties[key]
+
+    def __contains__(self, key: str) -> bool:
+        return key in self.additional_properties

bytekit/models/sitemap_response_sources_type_0.py

@@ -0,0 +1,44 @@
+from collections.abc import Mapping
+from typing import Any, TypeVar
+
+from attrs import define as _attrs_define
+from attrs import field as _attrs_field
+
+T = TypeVar("T", bound="SitemapResponseSourcesType0")
+
+
+@_attrs_define
+class SitemapResponseSourcesType0:
+    """ """
+
+    additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
+
+    def to_dict(self) -> dict[str, Any]:
+        field_dict: dict[str, Any] = {}
+        field_dict.update(self.additional_properties)
+
+        return field_dict
+
+    @classmethod
+    def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
+        d = dict(src_dict)
+        sitemap_response_sources_type_0 = cls()
+
+        sitemap_response_sources_type_0.additional_properties = d
+        return sitemap_response_sources_type_0
+
+    @property
+    def additional_keys(self) -> list[str]:
+        return list(self.additional_properties.keys())
+
+    def __getitem__(self, key: str) -> Any:
+        return self.additional_properties[key]
+
+    def __setitem__(self, key: str, value: Any) -> None:
+        self.additional_properties[key] = value
+
+    def __delitem__(self, key: str) -> None:
+        del self.additional_properties[key]
+
+    def __contains__(self, key: str) -> bool:
+        return key in self.additional_properties

bytekit/models/sitemap_response_status.py

@@ -0,0 +1,11 @@
+from enum import Enum
+
+
+class SitemapResponseStatus(str, Enum):
+    COMPLETED = "completed"
+    FAILED = "failed"
+    PENDING = "pending"
+    PROCESSING = "processing"
+
+    def __str__(self) -> str:
+        return str(self.value)

bytekit/models/viewport.py

@@ -0,0 +1,67 @@
+from collections.abc import Mapping
+from typing import Any, TypeVar
+
+from attrs import define as _attrs_define
+from attrs import field as _attrs_field
+
+T = TypeVar("T", bound="Viewport")
+
+
+@_attrs_define
+class Viewport:
+    """
+    Attributes:
+        width (int):
+        height (int):
+    """
+
+    width: int
+    height: int
+    additional_properties: dict[str, Any] = _attrs_field(init=False, factory=dict)
+
+    def to_dict(self) -> dict[str, Any]:
+        width = self.width
+
+        height = self.height
+
+        field_dict: dict[str, Any] = {}
+        field_dict.update(self.additional_properties)
+        field_dict.update(
+            {
+                "width": width,
+                "height": height,
+            }
+        )
+
+        return field_dict
+
+    @classmethod
+    def from_dict(cls: type[T], src_dict: Mapping[str, Any]) -> T:
+        d = dict(src_dict)
+        width = d.pop("width")
+
+        height = d.pop("height")
+
+        viewport = cls(
+            width=width,
+            height=height,
+        )
+
+        viewport.additional_properties = d
+        return viewport
+
+    @property
+    def additional_keys(self) -> list[str]:
+        return list(self.additional_properties.keys())
+
+    def __getitem__(self, key: str) -> Any:
+        return self.additional_properties[key]
+
+    def __setitem__(self, key: str, value: Any) -> None:
+        self.additional_properties[key] = value
+
+    def __delitem__(self, key: str) -> None:
+        del self.additional_properties[key]
+
+    def __contains__(self, key: str) -> bool:
+        return key in self.additional_properties