From 07d85c0bfc82b5df7366d3810c95aa8b959837a4 Mon Sep 17 00:00:00 2001 From: nytpu Date: Tue, 16 Mar 2021 22:43:18 -0600 Subject: implement New* routines for fetching and creating new feeds/pages. Also includes routines for fetching gemini, http, and gopher resources. TODO: add Update* routines to update existing feeds/pages. --- fetch/gemini.go | 133 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 133 insertions(+) create mode 100644 fetch/gemini.go (limited to 'fetch/gemini.go') diff --git a/fetch/gemini.go b/fetch/gemini.go new file mode 100644 index 0000000..d32d093 --- /dev/null +++ b/fetch/gemini.go @@ -0,0 +1,133 @@ +// Copyright (C) 2021 nytpu +// SPDX-License-Identifier: AGPL-3.0-or-later +// For more license details, see LICENSE or . + +// This file incorporates code from Drew DeVault's gemreader: +// https://git.sr.ht/~sircmpwn/gemreader/tree/master/item/feeds/fetch.go +// Original source is Copyright 2021 Drew DeVault and is licensed under the +// terms of the GNU Affero General Public License Version 3.0. +// https://www.gnu.org/licenses/agpl-3.0-standalone.html + +package fetch + +import ( + "context" + "crypto/sha256" + "fmt" + "io" + "mime" + "net/url" + "strings" + "time" + + "git.sr.ht/~adnano/go-gemini" + "github.com/mmcdole/gofeed" + + "golang.nytpu.com/comitium/core" +) + +// fetchGemini will make a request for a gemini resource +func fetchGemini(remoteURL *url.URL) (*gemini.Response, error) { + client := &gemini.Client{} + resp, err := client.Get(context.Background(), remoteURL.String()) + if err != nil { + return nil, err + } + if resp.Status != gemini.StatusSuccess { + return nil, fmt.Errorf("Unexpected Gemini response: %d %s", resp.Status, resp.Meta) + } + + return resp, nil +} + +// newGeminiFeed will fetch and create a new core.Feed given a gemini url +func newGeminiFeed(remoteURL *url.URL, title string) (*core.Feed, error) { + resp, err := fetchGemini(remoteURL) + if err != nil { + return nil, err + } + defer resp.Body.Close() + + mimetype, _, err := mime.ParseMediaType(resp.Meta) + if err != nil { + return nil, fmt.Errorf("Unintelligible content type: %s", resp.Meta) + } + + reader := io.LimitReader(resp.Body, 1073741824) // 1 GiB + + switch mimetype { + case "text/gemini": + var feed core.Feed + feed.Title = title // will be set later if given title is "" + feed.Link = remoteURL + text, err := gemini.ParseText(resp.Body) + if err != nil { + return nil, err + } + for _, line := range text { + switch line := line.(type) { + case gemini.LineHeading1: + if feed.Title == "" { + feed.Title = strings.TrimLeft(line.String(), "# ") + } + case gemini.LineLink: + if line.Name == "" || len(line.Name) < 10 { + continue + } + name := strings.TrimLeft(line.Name[:10], " -—") + date, err := time.Parse("2006-01-02", name) + if err != nil { + continue + } + link, err := url.Parse(line.URL) + if err != nil { + continue + } + link = remoteURL.ResolveReference(link) + item := &core.Item{} + item.Title = strings.TrimLeft(line.Name[10:], ": ") + item.Published = date + item.Link = link + feed.Items = append(feed.Items, item) + } + } + return &feed, nil + case "text/xml", + "application/rss+xml", + "application/atom+xml", + "application/xml", + "application/json": + gofeed, err := gofeed.NewParser().Parse(reader) + if err != nil { + return nil, err + } + feed, err := core.GofeedConvert(gofeed, title) + return feed, err + default: + return nil, fmt.Errorf("Cannot interpret %s as a feed", resp.Meta) + } +} + +// newGeminiPage will create a new core.Page given a gemini url +func newGeminiPage(remoteURL *url.URL, title string) (*core.Page, error) { + resp, err := fetchGemini(remoteURL) + if err != nil { + return nil, err + } + defer resp.Body.Close() + + reader := io.LimitReader(resp.Body, 1073741824) // 1 GiB + + var page core.Page + page.Title = title + page.Link = remoteURL + page.Updated = time.Now() + + h := sha256.New() + if _, err := io.Copy(h, reader); err != nil { + return nil, err + } + page.Hash = fmt.Sprintf("%x", h.Sum(nil)) + + return &page, nil +} -- cgit v1.2.3