Stop text overflowing the history card

frontend/eslint.config.mjs

@@ -33,12 +33,7 @@ export default [
 			"@typescript-eslint/class-methods-use-this": "off",
 			"@typescript-eslint/consistent-return": "off",
 			"@typescript-eslint/no-unsafe-argument": "off",
-			"@typescript-eslint/max-params": [
-				"error",
-				{
-					max: 6
-				}
-			],
+			"@typescript-eslint/max-params": "off",
 			"@typescript-eslint/no-magic-numbers": "off",
 			"@typescript-eslint/prefer-readonly-parameter-types": "off",
 			"@typescript-eslint/naming-convention": "off",

frontend/package-lock.json

@@ -1583,7 +1583,9 @@
 			}
 		},
 		"node_modules/brace-expansion": {
-			"version": "1.1.11",
+			"version": "1.1.12",
+			"resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-1.1.12.tgz",
+			"integrity": "sha512-9T9UjW3r0UW5c1Q7GTwllptXwhvYmEzFhzMfZ9H7FQWt+uZePjZPjBP/W1ZEyZ1twGWom5/56TF4lPcqjnDHcg==",
 			"dev": true,
 			"license": "MIT",
 			"dependencies": {
@@ -2742,7 +2744,9 @@
 			}
 		},
 		"node_modules/js-yaml": {
-			"version": "4.1.0",
+			"version": "4.1.1",
+			"resolved": "https://registry.npmjs.org/js-yaml/-/js-yaml-4.1.1.tgz",
+			"integrity": "sha512-qQKT4zQxXl8lLwBtHMWwaTcGfFOZviOJet3Oy/xmGk2gZH677CJM9EvtfdSkgWcATZhj/55JZ0rmy3myCT5lsA==",
 			"dev": true,
 			"license": "MIT",
 			"dependencies": {
@@ -3487,6 +3491,7 @@
 			"version": "0.5.0",
 			"resolved": "https://registry.npmjs.org/reconcile-text/-/reconcile-text-0.5.0.tgz",
 			"integrity": "sha512-zki3lqw9Oxdhm9ZvDN17VyYoL1Isc8BEL07ILVDE2yGfNEI7thrkczoNCUr+hkFU2rzZtfxECTG0b7p61AJ6wg==",
+			"dev": true,
 			"license": "MIT"
 		},
 		"node_modules/regex-parser": {
@@ -4687,7 +4692,7 @@
 				"byte-base64": "^1.1.0",
 				"minimatch": "^10.0.1",
 				"p-queue": "^8.1.0",
-				"reconcile-text": "^0.5.0",
+				"reconcile-text": "^0.7.1",
 				"uuid": "^13.0.0"
 			},
 			"devDependencies": {
@@ -4703,7 +4708,9 @@
 			}
 		},
 		"sync-client/node_modules/brace-expansion": {
-			"version": "2.0.1",
+			"version": "2.0.2",
+			"resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.2.tgz",
+			"integrity": "sha512-Jt0vHyM+jmUBqojB7E1NIYadt0vI0Qxjxd2TErW94wDz+E2LAm5vKMXXwg6ZZBTHPuUlDgQHKXvjGBdfcF1ZDQ==",
 			"license": "MIT",
 			"dependencies": {
 				"balanced-match": "^1.0.0"
@@ -4722,6 +4729,12 @@
 				"url": "https://github.com/sponsors/isaacs"
 			}
 		},
+		"sync-client/node_modules/reconcile-text": {
+			"version": "0.7.1",
+			"resolved": "https://registry.npmjs.org/reconcile-text/-/reconcile-text-0.7.1.tgz",
+			"integrity": "sha512-khedcYvAKs7ELKh5Z8mz2vyomMY5TqznV1dB+k/7qUAX9cheMNN5/EPJVQYZepOMunYbnQitvhFJX3kD4IMcNw==",
+			"license": "MIT"
+		},
 		"test-client": {
 			"version": "0.9.2",
 			"bin": {

frontend/sync-client/package.json

@@ -16,7 +16,7 @@
         "byte-base64": "^1.1.0",
         "minimatch": "^10.0.1",
         "p-queue": "^8.1.0",
-        "reconcile-text": "^0.5.0",
+        "reconcile-text": "^0.7.1",
         "uuid": "^13.0.0"
     },
     "devDependencies": {

frontend/sync-client/src/file-operations/file-operations.ts

@@ -3,8 +3,9 @@ import type { FileSystemOperations } from "./filesystem-operations";
 import type { Database, RelativePath } from "../persistence/database";
 import { SafeFileSystemOperations } from "./safe-filesystem-operations";
 import type { TextWithCursors } from "reconcile-text";
-import { isBinary, reconcile } from "reconcile-text";
+import { reconcile } from "reconcile-text";
 import { isFileTypeMergable } from "../utils/is-file-type-mergable";
+import { isBinary } from "../utils/is-binary";
 
 export class FileOperations {
 	private static readonly PARENTHESES_REGEX = / \((\d+)\)$/;

frontend/sync-client/src/services/sync-service.ts

@@ -16,6 +16,7 @@ import type { DocumentVersion } from "./types/DocumentVersion";
 import type { FetchLatestDocumentsResponse } from "./types/FetchLatestDocumentsResponse";
 import type { PingResponse } from "./types/PingResponse";
 import type { DeleteDocumentVersion } from "./types/DeleteDocumentVersion";
+import type { UpdateTextDocumentVersion } from "./types/UpdateTextDocumentVersion";
 
 export interface CheckConnectionResult {
 	isSuccessful: boolean;
@@ -102,7 +103,64 @@ export class SyncService {
 		});
 	}
 
-	public async put({
+	public async putText({
+		parentVersionId,
+		documentId,
+		relativePath,
+		content
+	}: {
+		parentVersionId: VaultUpdateId;
+		documentId: DocumentId;
+		relativePath: RelativePath;
+		content: (number | string | bigint)[];
+	}): Promise<DocumentUpdateResponse> {
+		return this.withRetries(async () => {
+			this.logger.debug(
+				`Updating text document ${documentId} with parent version ${parentVersionId} and relative path ${relativePath}`
+			);
+
+			const request: UpdateTextDocumentVersion = {
+				parentVersionId,
+				relativePath,
+				content: content.map((c) => {
+					if (typeof c === "bigint") {
+						return Number(c);
+					}
+					return c;
+				})
+			};
+
+			const response = await this.client(
+				this.getUrl(`/documents/${documentId}/text`),
+				{
+					method: "PUT",
+					body: JSON.stringify(request),
+					headers: this.getDefaultHeaders({ type: "json" })
+				}
+			);
+
+			const result: SerializedError | DocumentUpdateResponse =
+				(await response.json()) as  // eslint-disable-line @typescript-eslint/no-unsafe-type-assertion
+					| SerializedError
+					| DocumentUpdateResponse;
+
+			if ("errorType" in result) {
+				throw new Error(
+					`Failed to update document: ${SyncService.formatError(result)}`
+				);
+			}
+
+			this.logger.debug(
+				`Updated document ${JSON.stringify(result)} with id ${
+					result.documentId
+				}}`
+			);
+
+			return result;
+		});
+	}
+
+	public async putBinary({
 		parentVersionId,
 		documentId,
 		relativePath,
@@ -115,7 +173,7 @@ export class SyncService {
 	}): Promise<DocumentUpdateResponse> {
 		return this.withRetries(async () => {
 			this.logger.debug(
-				`Updating document ${documentId} with parent version ${parentVersionId} and relative path ${relativePath}`
+				`Updating binary document ${documentId} with parent version ${parentVersionId} and relative path ${relativePath}`
 			);
 			const formData = new FormData();
 			formData.append("parent_version_id", parentVersionId.toString());
@@ -126,7 +184,7 @@ export class SyncService {
 			);
 
 			const response = await this.client(
-				this.getUrl(`/documents/${documentId}`),
+				this.getUrl(`/documents/${documentId}/binary`),
 				{
 					method: "PUT",
 					body: formData,
@@ -171,10 +229,7 @@ export class SyncService {
 				{
 					method: "DELETE",
 					body: JSON.stringify(request),
-					headers: {
-						"Content-Type": "application/json",
-						...this.getDefaultHeaders()
-					}
+					headers: this.getDefaultHeaders({ type: "json" })
 				}
 			);
 
@@ -297,11 +352,21 @@ export class SyncService {
 		return `${safeRemoteUri}/vaults/${vaultName}${path}`;
 	}
 
-	private getDefaultHeaders(): Record<string, string> {
-		return {
+	private getDefaultHeaders(
+		{ type }: { type?: "json" | "form" } = { type: undefined }
+	): Record<string, string> {
+		const headers: Record<string, string> = {
 			"device-id": this.deviceId,
 			authorization: `Bearer ${this.settings.getSettings().token}`
 		};
+
+		if (type === "json") {
+			headers["Content-Type"] = "application/json";
+		} else if (type === "form") {
+			headers["Content-Type"] = "multipart/form-data";
+		}
+
+		return headers;
 	}
 
 	private async withRetries<T>(fn: () => Promise<T>): Promise<T> {

frontend/sync-client/src/services/types/UpdateBinaryDocumentVersion.ts

@@ -0,0 +1,7 @@
+// This file was generated by [ts-rs](https://github.com/Aleph-Alpha/ts-rs). Do not edit this file manually.
+
+export interface UpdateBinaryDocumentVersion {
+	parent_version_id: bigint;
+	relative_path: string;
+	content: number[];
+}

frontend/sync-client/src/services/types/UpdateTextDocumentVersion.ts

@@ -0,0 +1,7 @@
+// This file was generated by [ts-rs](https://github.com/Aleph-Alpha/ts-rs). Do not edit this file manually.
+
+export interface UpdateTextDocumentVersion {
+	parentVersionId: number;
+	relativePath: string;
+	content: (number | string)[];
+}

frontend/sync-client/src/sync-client.ts

@@ -21,13 +21,13 @@ import { CursorTracker } from "./sync-operations/cursor-tracker";
 import type { CursorSpan } from "./services/types/CursorSpan";
 import type { MaybeOutdatedClientCursors } from "./types/maybe-outdated-client-cursors";
 import { FileChangeNotifier } from "./sync-operations/file-change-notifier";
+import { FixedSizeDocumentCache } from "./utils/fix-sized-cache";
 
 export class SyncClient {
 	private static readonly MINIMUM_SAVE_INTERVAL_MS = 1000;
 	private hasStartedOfflineSync = false;
 	private hasFinishedOfflineSync = false;
 
-	// eslint-disable-next-line @typescript-eslint/max-params
 	private constructor(
 		private readonly history: SyncHistory,
 		private readonly settings: Settings,
@@ -135,13 +135,15 @@ export class SyncClient {
 			nativeLineEndings
 		);
 
+		const contentCache = new FixedSizeDocumentCache(1024 * 1024 * 2); // 2 MB cache
 		const unrestrictedSyncer = new UnrestrictedSyncer(
 			logger,
 			database,
 			settings,
 			syncService,
 			fileOperations,
-			history
+			history,
+			contentCache
 		);
 
 		const syncer = new Syncer(

frontend/sync-client/src/sync-operations/unrestricted-syncer.ts

@@ -4,6 +4,7 @@ import type {
 	RelativePath
 } from "../persistence/database";
 
+import { diff } from "reconcile-text";
 import type { SyncService } from "../services/sync-service";
 import type { Logger } from "../tracing/logger";
 import type {
@@ -27,6 +28,9 @@ import { globsToRegexes } from "../utils/globs-to-regexes";
 import type { DocumentVersion } from "../services/types/DocumentVersion";
 import type { DocumentUpdateResponse } from "../services/types/DocumentUpdateResponse";
 import type { DocumentVersionWithoutContent } from "../services/types/DocumentVersionWithoutContent";
+import type { FixedSizeDocumentCache } from "../utils/fix-sized-cache";
+import { isFileTypeMergable } from "../utils/is-file-type-mergable";
+import { isBinary } from "../utils/is-binary";
 
 export class UnrestrictedSyncer {
 	private ignorePatterns: RegExp[];
@@ -37,7 +41,8 @@ export class UnrestrictedSyncer {
 		private readonly settings: Settings,
 		private readonly syncService: SyncService,
 		private readonly operations: FileOperations,
-		private readonly history: SyncHistory
+		private readonly history: SyncHistory,
+		private readonly contentCache: FixedSizeDocumentCache
 	) {
 		this.ignorePatterns = globsToRegexes(
 			this.settings.getSettings().ignorePatterns,
@@ -87,8 +92,12 @@ export class UnrestrictedSyncer {
 				},
 				document
 			);
-
 			this.database.addSeenUpdateId(response.vaultUpdateId);
+			this.updateCache(
+				response.vaultUpdateId,
+				contentBytes,
+				response.relativePath
+			);
 
 			this.history.addHistoryEntry({
 				status: SyncStatus.SUCCESS,
@@ -178,12 +187,32 @@ export class UnrestrictedSyncer {
 				undefined;
 
 			if (areThereLocalChanges) {
-				response = await this.syncService.put({
-					documentId: document.documentId,
-					parentVersionId: document.metadata.parentVersionId,
-					relativePath: document.relativePath,
-					contentBytes
-				});
+				const isText =
+					!isBinary(contentBytes) &&
+					isFileTypeMergable(document.relativePath);
+				const cachedVersion = this.contentCache.get(
+					document.metadata.parentVersionId
+				);
+
+				response =
+					isText && cachedVersion !== undefined
+						? await this.syncService.putText({
+								documentId: document.documentId,
+								parentVersionId:
+									document.metadata.parentVersionId,
+								relativePath: document.relativePath,
+								content: diff(
+									new TextDecoder().decode(cachedVersion),
+									new TextDecoder().decode(contentBytes)
+								)
+							})
+						: await this.syncService.putBinary({
+								documentId: document.documentId,
+								parentVersionId:
+									document.metadata.parentVersionId,
+								relativePath: document.relativePath,
+								contentBytes
+							});
 			} else {
 				if (!force) {
 					this.logger.debug(
@@ -274,12 +303,16 @@ export class UnrestrictedSyncer {
 					},
 					document
 				);
-
 				await this.operations.write(
 					actualPath,
 					contentBytes,
 					responseBytes
 				);
+				this.updateCache(
+					response.vaultUpdateId,
+					contentBytes,
+					actualPath
+				);
 
 				if (!force) {
 					this.history.addHistoryEntry({
@@ -297,6 +330,11 @@ export class UnrestrictedSyncer {
 					},
 					document
 				);
+				this.updateCache(
+					response.vaultUpdateId,
+					contentBytes,
+					actualPath
+				);
 			}
 
 			this.database.addSeenUpdateId(response.vaultUpdateId);
@@ -423,6 +461,11 @@ export class UnrestrictedSyncer {
 				remoteVersion.relativePath,
 				contentBytes
 			);
+			this.updateCache(
+				remoteVersion.vaultUpdateId,
+				contentBytes,
+				remoteVersion.relativePath
+			);
 
 			resolve();
 			this.database.removeDocumentPromise(promise);
@@ -513,4 +556,14 @@ export class UnrestrictedSyncer {
 			};
 		}
 	}
+
+	private updateCache(
+		updateId: number,
+		contentBytes: Uint8Array,
+		filePath: RelativePath
+	): void {
+		if (isFileTypeMergable(filePath) && !isBinary(contentBytes)) {
+			this.contentCache.put(updateId, contentBytes);
+		}
+	}
 }

frontend/sync-client/src/utils/fix-sized-cache.test.ts

@@ -0,0 +1,62 @@
+import { describe, it } from "node:test";
+import assert from "node:assert";
+import { FixedSizeDocumentCache } from "./fix-sized-cache";
+
+describe("fixedSizeDocumentCache", () => {
+	it("happyPath", async () => {
+		const cache = new FixedSizeDocumentCache(4);
+		const doc1 = new Uint8Array([1, 2]);
+		const doc2 = new Uint8Array([3, 4]);
+		const doc3 = new Uint8Array([5, 6]);
+
+		cache.put(1, doc1);
+		assert.equal(cache.get(1), doc1);
+		cache.put(2, doc2);
+		assert.equal(cache.get(1), doc1);
+		assert.equal(cache.get(2), doc2);
+		cache.put(3, doc3);
+		assert.equal(cache.get(1), undefined);
+		assert.equal(cache.get(2), doc2);
+		assert.equal(cache.get(3), doc3);
+	});
+
+	it("updateExistingEntry", async () => {
+		const cache = new FixedSizeDocumentCache(4);
+		const doc1_v1 = new Uint8Array([1, 2]);
+		const doc1_v2 = new Uint8Array([3, 4]);
+		const doc2 = new Uint8Array([5, 6]);
+
+		cache.put(1, doc1_v1);
+		assert.equal(cache.get(1), doc1_v1);
+		cache.put(2, doc2);
+		assert.equal(cache.get(1), doc1_v1);
+		assert.equal(cache.get(2), doc2);
+		cache.put(1, doc1_v2); // Update doc1
+		assert.equal(cache.get(1), doc1_v2);
+		assert.equal(cache.get(2), doc2);
+	});
+
+	it("evictOldestEntry", async () => {
+		const cache = new FixedSizeDocumentCache(4);
+		const doc1 = new Uint8Array([1, 2]);
+		const doc2 = new Uint8Array([3, 4]);
+		const doc3 = new Uint8Array([5, 6]);
+
+		cache.put(1, doc1);
+		cache.put(2, doc2);
+		assert.equal(cache.get(2), doc2);
+		assert.equal(cache.get(1), doc1);
+		cache.put(3, doc3);
+		assert.equal(cache.get(1), doc1);
+		assert.equal(cache.get(2), undefined);
+		assert.equal(cache.get(3), doc3);
+	});
+
+	it("tooLargeEntry", async () => {
+		const cache = new FixedSizeDocumentCache(2);
+		const doc1 = new Uint8Array([1, 2, 3]);
+
+		cache.put(1, doc1);
+		assert.equal(cache.get(1), undefined);
+	});
+});

frontend/sync-client/src/utils/fix-sized-cache.ts

@@ -0,0 +1,57 @@
+// Implements an in-memory fixed-size cache for document contents,
+
+import type { VaultUpdateId } from "../persistence/database";
+
+// evicting the least recently used documents when the size limit is exceeded.
+export class FixedSizeDocumentCache {
+	private readonly maxSizeInBytes: number;
+	private currentSizeInBytes: number;
+	private readonly cache: Map<VaultUpdateId, Uint8Array>;
+	private usageOrder: VaultUpdateId[];
+
+	public constructor(maxSizeInBytes: number) {
+		this.maxSizeInBytes = maxSizeInBytes;
+		this.currentSizeInBytes = 0;
+		this.cache = new Map();
+		this.usageOrder = [];
+	}
+
+	public get(updateId: VaultUpdateId): Uint8Array | undefined {
+		const entry = this.cache.get(updateId);
+		if (entry) {
+			this.usageOrder = this.usageOrder.filter((id) => id !== updateId);
+			this.usageOrder.push(updateId);
+			return entry;
+		}
+		return undefined;
+	}
+
+	public put(updateId: VaultUpdateId, content: Uint8Array): void {
+		if (content.byteLength > this.maxSizeInBytes) {
+			// Document is too large to fit in the cache
+			return;
+		}
+
+		// If the document is already in the cache, update it
+		const existingEntry = this.cache.get(updateId);
+		if (existingEntry != null) {
+			this.currentSizeInBytes -= existingEntry.byteLength;
+			this.cache.delete(updateId);
+			this.usageOrder = this.usageOrder.filter((id) => id !== updateId);
+		}
+		this.cache.set(updateId, content);
+		this.usageOrder.push(updateId);
+		this.currentSizeInBytes += content.byteLength;
+
+		// Evict least recently used documents if over size limit
+		while (
+			this.currentSizeInBytes > this.maxSizeInBytes &&
+			this.usageOrder.length > 0
+		) {
+			const lruUpdateId = this.usageOrder.shift()!; // eslint-disable-line @typescript-eslint/no-non-null-assertion
+			const lruEntry = this.cache.get(lruUpdateId)!; // eslint-disable-line @typescript-eslint/no-non-null-assertion
+			this.cache.delete(lruUpdateId);
+			this.currentSizeInBytes -= lruEntry.byteLength;
+		}
+	}
+}

frontend/sync-client/src/utils/is-binary.ts

@@ -0,0 +1,16 @@
+// Text is unlikely to contain null bytes, so we can use that to distinguish binary files.
+export function isBinary(content: Uint8Array): boolean {
+	for (const byte of content) {
+		if (byte === 0) {
+			return true;
+		}
+	}
+
+	try {
+		new TextDecoder("utf-8", { fatal: true }).decode(content);
+	} catch {
+		return true;
+	}
+
+	return false;
+}

sync-server/Cargo.lock

@@ -1680,9 +1680,12 @@ dependencies = [
 
 [[package]]
 name = "reconcile-text"
-version = "0.5.0"
+version = "0.7.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c8d690c19b0bf6574cd3591d10f20df5aa52d2af95b8dcaacbc86893292ac8c5"
+checksum = "913440a3c2b90cd3ed3e967660f2bb624b71e8059b9fc86960a5f91bd1e2e353"
+dependencies = [
+ "serde",
+]
 
 [[package]]
 name = "redox_syscall"

sync-server/Cargo.toml

@@ -35,7 +35,7 @@ bimap = "0.6.3"
 ts-rs =  { version = "10.1", features = ["uuid-impl", "chrono-impl"] }
 serde_with = "3.15.1"
 base64 = "0.22.1"
-reconcile-text = "0.5.0"
+reconcile-text = { version = "0.7.1", features = ["serde"] }
 
 [profile.release]
 codegen-units = 1

sync-server/src/server.rs

@@ -117,8 +117,12 @@ fn get_authed_routes(app_state: AppState) -> Router<AppState> {
             get(fetch_latest_document_version::fetch_latest_document_version),
         )
         .route(
-            "/vaults/:vault_id/documents/:document_id",
-            put(update_document::update_document),
+            "/vaults/:vault_id/documents/:document_id/binary",
+            put(update_document::update_binary),
+        )
+        .route(
+            "/vaults/:vault_id/documents/:document_id/text",
+            put(update_document::update_text),
         )
         .route(
             "/vaults/:vault_id/documents/:document_id/versions/:version_id",

sync-server/src/server/requests.rs

@@ -1,5 +1,6 @@
 use axum::body::Bytes;
 use axum_typed_multipart::{FieldData, TryFromMultipart};
+use reconcile_text::NumberOrString;
 use serde::{self, Deserialize};
 use ts_rs::TS;
 
@@ -22,7 +23,7 @@ pub struct CreateDocumentVersion {
 
 #[derive(TS, Debug, TryFromMultipart)]
 #[ts(export)]
-pub struct UpdateDocumentVersion {
+pub struct UpdateBinaryDocumentVersion {
     pub parent_version_id: VaultUpdateId,
     pub relative_path: String,
 
@@ -31,6 +32,19 @@ pub struct UpdateDocumentVersion {
     pub content: FieldData<Bytes>,
 }
 
+#[derive(TS, Debug, Deserialize)]
+#[serde(rename_all = "camelCase")]
+#[ts(export)]
+pub struct UpdateTextDocumentVersion {
+    #[ts(as = "i32")]
+    pub parent_version_id: VaultUpdateId,
+
+    pub relative_path: String,
+
+    #[ts(type = "Array<number | string>")]
+    pub content: Vec<NumberOrString>,
+}
+
 #[derive(TS, Debug, Deserialize)]
 #[serde(rename_all = "camelCase")]
 #[ts(export)]

sync-server/src/server/update_document.rs

@@ -6,23 +6,25 @@ use axum::{
 use axum_extra::TypedHeader;
 use axum_typed_multipart::TypedMultipart;
 use log::info;
-use reconcile_text::{BuiltinTokenizer, is_binary, reconcile};
+use reconcile_text::{BuiltinTokenizer, EditedText, reconcile};
 use serde::Deserialize;
 
 use super::{
-    device_id_header::DeviceIdHeader, requests::UpdateDocumentVersion,
+    device_id_header::DeviceIdHeader, requests::UpdateTextDocumentVersion,
     responses::DocumentUpdateResponse,
 };
 use crate::{
     app_state::{
         AppState,
-        database::models::{DocumentId, StoredDocumentVersion, VaultId},
+        database::models::{DocumentId, StoredDocumentVersion, VaultId, VaultUpdateId},
     },
     config::user_config::User,
     errors::{SyncServerError, not_found_error, server_error},
+    server::requests::UpdateBinaryDocumentVersion,
     utils::{
-        dedup_paths::dedup_paths, is_file_type_mergable::is_file_type_mergable,
-        normalize::normalize, sanitize_path::sanitize_path,
+        dedup_paths::dedup_paths, is_binary::is_binary,
+        is_file_type_mergable::is_file_type_mergable, normalize::normalize,
+        sanitize_path::sanitize_path,
     },
 };
 
@@ -30,13 +32,11 @@ use crate::{
 pub struct UpdateDocumentPathParams {
     #[serde(deserialize_with = "normalize")]
     vault_id: VaultId,
-
     document_id: DocumentId,
 }
 
 #[axum::debug_handler]
-#[allow(clippy::too_many_lines)]
-pub async fn update_document(
+pub async fn update_binary(
     Path(UpdateDocumentPathParams {
         vault_id,
         document_id,
@@ -44,25 +44,92 @@ pub async fn update_document(
     Extension(user): Extension<User>,
     TypedHeader(device_id): TypedHeader<DeviceIdHeader>,
     State(state): State<AppState>,
-    TypedMultipart(request): TypedMultipart<UpdateDocumentVersion>,
+    TypedMultipart(request): TypedMultipart<UpdateBinaryDocumentVersion>,
 ) -> Result<Json<DocumentUpdateResponse>, SyncServerError> {
-    // No need for a transaction as document versions are immutable
-    let parent_document = state
+    let parent_document = get_parent_document(&state, &vault_id, request.parent_version_id).await?;
+    let content = request.content.contents.to_vec();
+
+    update_document(
+        parent_document,
+        vault_id,
+        document_id,
+        user,
+        device_id,
+        state,
+        &request.relative_path,
+        content,
+    )
+    .await
+}
+
+#[axum::debug_handler]
+#[allow(clippy::too_many_lines)]
+pub async fn update_text(
+    Path(UpdateDocumentPathParams {
+        vault_id,
+        document_id,
+    }): Path<UpdateDocumentPathParams>,
+    Extension(user): Extension<User>,
+    TypedHeader(device_id): TypedHeader<DeviceIdHeader>,
+    State(state): State<AppState>,
+    Json(request): Json<UpdateTextDocumentVersion>,
+) -> Result<Json<DocumentUpdateResponse>, SyncServerError> {
+    let parent_document = get_parent_document(&state, &vault_id, request.parent_version_id).await?;
+
+    let edited_text = EditedText::from_diff(
+        str::from_utf8(&parent_document.content)
+            .expect("parent must be valid UTF-8 because it's a text document"),
+        request.content,
+        &*BuiltinTokenizer::Word,
+    );
+
+    let content = edited_text.apply().text().into_bytes();
+
+    update_document(
+        parent_document,
+        vault_id,
+        document_id,
+        user,
+        device_id,
+        state,
+        &request.relative_path,
+        content,
+    )
+    .await
+}
+
+async fn get_parent_document(
+    state: &AppState,
+    vault_id: &VaultId,
+    parent_version_id: VaultUpdateId,
+) -> Result<StoredDocumentVersion, SyncServerError> {
+    state
         .database
-        .get_document_version(&vault_id, request.parent_version_id, None)
+        .get_document_version(vault_id, parent_version_id, None)
         .await
         .map_err(server_error)?
         .map_or_else(
             || {
                 Err(not_found_error(anyhow!(
-                    "Parent version with id `{}` not found",
-                    request.parent_version_id
+                    "Parent version with id `{parent_version_id}` not found"
                 )))
             },
             Ok,
-        )?;
+        )
+}
 
-    let sanitized_relative_path = sanitize_path(&request.relative_path);
+#[allow(clippy::too_many_lines, clippy::too_many_arguments)]
+async fn update_document(
+    parent_document: StoredDocumentVersion,
+    vault_id: VaultId,
+    document_id: DocumentId,
+    user: User,
+    device_id: DeviceIdHeader,
+    state: AppState,
+    relative_path: &str,
+    content: Vec<u8>,
+) -> Result<Json<DocumentUpdateResponse>, SyncServerError> {
+    let sanitized_relative_path = sanitize_path(relative_path);
 
     let mut transaction = state
         .database
@@ -102,8 +169,6 @@ pub async fn update_document(
         )));
     }
 
-    let content = request.content.contents.to_vec();
-
     // Return the latest version if the content and path are the same as the latest
     // version
     if content == latest_version.content && sanitized_relative_path == latest_version.relative_path

sync-server/src/utils.rs

@@ -1,4 +1,5 @@
 pub mod dedup_paths;
+pub mod is_binary;
 pub mod is_file_type_mergable;
 pub mod normalize;
 pub mod rotating_file_writer;

sync-server/src/utils/is_binary.rs

@@ -0,0 +1,26 @@
+/// Heuristically determine if the given data is a binary or a text file's
+/// content.
+///
+/// Only text inputs can be reconciled using the crate's functions.
+#[must_use]
+pub fn is_binary(data: &[u8]) -> bool {
+    if data.contains(&0) {
+        // Even though the NUL character is valid in UTF-8, it's highly suspicious in
+        // human-readable text.
+        return true;
+    }
+
+    std::str::from_utf8(data).is_err()
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_is_binary() {
+        assert!(is_binary(&[0, 159, 146, 150]));
+        assert!(is_binary(&[0, 12]));
+        assert!(!is_binary(b"hello"));
+    }
+}