482a46b568
Updated project files for `Beam.Dynamic`, `Beam.Exports`, `Beam.Temporary.Cli`, and `Beam` to include additional metadata and specific package versions. Refactored `DataBindings` and `ResolvedBindings` to records, added a new `Text` property in `Binding.cs`, and introduced `ParseNumbers` in `OnlineCleaner`. New classes `PuppetContext` and `PuppetUnitDownloader` added for Playwright integration. Introduced `ImmutableState` struct and `UnitDownloaderBinary` class for improved download management. Updated tests in `UnitTest1.cs` for number localization. Added `Beam.Puppeteer` project to the solution.
174 lines
7.8 KiB
C#
174 lines
7.8 KiB
C#
using aeqw89.DataKeys;
|
||
using Beam.Dynamic;
|
||
using Beam;
|
||
using Microsoft.Extensions.Logging;
|
||
using System;
|
||
using System.Collections.Generic;
|
||
|
||
namespace Beam.Temporary.Cli {
|
||
/// <summary>
|
||
/// Type‑safe, staged builder that prevents callers from forgetting the mandatory steps
|
||
/// (source → link selection → transformer) and surfaces operational knobs as first‑class
|
||
/// methods instead of magic parameters.
|
||
/// </summary>
|
||
public static class DownloadBuilder<T> {
|
||
/* ──────────────────────────── Entry points ─────────────────────────── */
|
||
|
||
public static ILinkStage FromMeta(DataKey<TextResource> novelKey, BeamDataDictionary data) =>
|
||
Create(novelKey, data, SourceKind.Meta);
|
||
|
||
public static ILinkStage FromText(DataKey<TextResource> novelKey, BeamDataDictionary data) =>
|
||
Create(novelKey, data, SourceKind.Text);
|
||
|
||
/* ────────────────────────────── Stages ─────────────────────────────── */
|
||
|
||
public interface ILinkStage {
|
||
ITransformStage WithLink();
|
||
ITransformStage WithLinkGenerator();
|
||
ILinkStage WithRange(Range range);
|
||
}
|
||
|
||
public interface ITransformStage {
|
||
IContextStage<U> WithTransformer<U>(Func<DataBindings, AsyncTransformer<T, U>> factory);
|
||
}
|
||
|
||
public interface IContextStage<U> {
|
||
IContextStage<U> Configure(Action<DownloadContextBuilder<T>> configure);
|
||
IContextStage<U> WithParallelism(int degree);
|
||
IContextStage<U> WithTimeout(TimeSpan timeout);
|
||
IContextStage<U> WithRetryReporter(IProgress<RetryReport> reporter);
|
||
DownloadEnumerable<T> Build();
|
||
}
|
||
|
||
/* ────────────────────────── Implementation ────────────────────────── */
|
||
|
||
private enum SourceKind { Meta, Text }
|
||
|
||
private static ILinkStage Create(DataKey<TextResource> novelKey, BeamDataDictionary data, SourceKind kind) {
|
||
var (source, initial) = Resolve(novelKey, data, kind);
|
||
var ctxBuilder = new DownloadContextBuilder<T>().WithLinks(Array.Empty<SourceLink>()); // placeholder, filled later.
|
||
return new LinkStage(source, initial, data, ctxBuilder);
|
||
}
|
||
|
||
private static (WebResource Source, State Initial) Resolve(DataKey<TextResource> novelKey, BeamDataDictionary data, SourceKind kind) {
|
||
if (!data.Novels.TryGetValue(novelKey, out var tr))
|
||
throw new KeyNotFoundException($"Novel '{novelKey}' not found in BeamDataDictionary.");
|
||
|
||
var textRecord = tr.ToRecord(data);
|
||
WebResource? source;
|
||
State? initial;
|
||
|
||
if (kind == SourceKind.Meta) {
|
||
source = textRecord.AssociatedMetaSource ?? throw new InvalidOperationException($"Meta source missing for '{novelKey}'.");
|
||
initial = textRecord.Resource.MetaTemplateInitialData ?? throw new InvalidOperationException("Meta template data missing.");
|
||
} else {
|
||
source = textRecord.AssociatedSource ?? throw new InvalidOperationException($"Text source missing for '{novelKey}'.");
|
||
initial = textRecord.Resource.TemplateInitialData;
|
||
}
|
||
|
||
return (source, initial);
|
||
}
|
||
|
||
/* ──────────────────────────── Stage types ─────────────────────────── */
|
||
|
||
private sealed record LinkStage(
|
||
WebResource Source,
|
||
State Initial,
|
||
BeamDataDictionary Data,
|
||
DownloadContextBuilder<T> CtxBuilder) : ILinkStage {
|
||
|
||
private State? endState;
|
||
private bool linksFrozen = false;
|
||
|
||
public ITransformStage WithLink() {
|
||
var link = Data.Templates[Source.Key].Builder.Build(Initial);
|
||
CtxBuilder.WithLinks(new[] { link });
|
||
return new TransformStage(Source, Data, CtxBuilder);
|
||
}
|
||
|
||
public ITransformStage WithLinkGenerator() {
|
||
var template = Data.Templates[Source.Key];
|
||
var generator = SourceLinkEnumerable.FromGenerator(new OrderedSourceLinkGenerator(
|
||
template.Builder,
|
||
new NumberedStateChanger(template.Factory.Behavior),
|
||
Initial, endState));
|
||
CtxBuilder.WithLinks(generator);
|
||
linksFrozen = true;
|
||
return new TransformStage(Source, Data, CtxBuilder);
|
||
}
|
||
|
||
public ILinkStage WithRange(Range range) {
|
||
if (linksFrozen)
|
||
throw new InvalidOperationException($"WithRange must be called before WithLinkGenerator");
|
||
if (range.End.Value < range.Start.Value)
|
||
throw new ArgumentOutOfRangeException(nameof(range), $" start must be < end");
|
||
var template = Data.Templates[Source.Key];
|
||
var stateChanger = new NumberedStateChanger(template.Factory.Behavior);
|
||
endState = Initial.Copy();
|
||
stateChanger.Apply(Initial, range.Start.Value - 1);
|
||
stateChanger.Apply(endState, range.End.Value - 1);
|
||
return this;
|
||
}
|
||
}
|
||
|
||
private sealed record TransformStage(
|
||
WebResource Source,
|
||
BeamDataDictionary Data,
|
||
DownloadContextBuilder<T> CtxBuilder) : ITransformStage {
|
||
public IContextStage WithTransformer<U>(Func<DataBindings, Func<object, T>> factory) {
|
||
var transformer = factory(Data.Bindings[Source.Bindings]);
|
||
return new ContextStage<U>(CtxBuilder, transformer);
|
||
}
|
||
}
|
||
|
||
private sealed class ContextStage<U> : IContextStage {
|
||
private readonly DownloadContextBuilder<T> _ctxBuilder;
|
||
private readonly Func<object, T> _transformer;
|
||
private int _parallelism = 4;
|
||
|
||
public ContextStage(DownloadContextBuilder<T> ctxBuilder, Func<object, T> transformer) {
|
||
_ctxBuilder = ctxBuilder;
|
||
_transformer = transformer;
|
||
}
|
||
|
||
public IContextStage Configure(Action<DownloadContextBuilder<T>> configure) {
|
||
configure(_ctxBuilder);
|
||
return this;
|
||
}
|
||
|
||
public IContextStage WithParallelism(int degree) {
|
||
_parallelism = Math.Max(1, degree);
|
||
return this;
|
||
}
|
||
|
||
public IContextStage WithTimeout(TimeSpan timeout) {
|
||
_ctxBuilder.WithTimeOut(timeout);
|
||
return this;
|
||
}
|
||
|
||
public IContextStage WithRetryReporter(IProgress<RetryReport> reporter) {
|
||
_ctxBuilder.WithRetryReporter(reporter);
|
||
return this;
|
||
}
|
||
|
||
public DownloadEnumerable<T> Build() {
|
||
var context = _ctxBuilder.Build();
|
||
SequentialFragmentDownloader<T> sequentialDownloader = new(
|
||
context,
|
||
ctx => new UnitFragmentDownloader<T>(
|
||
context.Web,
|
||
_transformer,
|
||
context.AsyncFailurePredicates,
|
||
_parallelism,
|
||
context.DownloadLogger),
|
||
context.DownloadLogger);
|
||
var enumerable = new DownloadEnumerable<T>(
|
||
sequentialDownloader
|
||
.UnwrapFragmented());
|
||
sequentialDownloader.DisposeAsync().AsTask().Wait();
|
||
return enumerable;
|
||
}
|
||
}
|
||
}
|
||
}
|