NDF
中文版
NDF语言是一种用于描述游戏数据的脚本语言。
[!NOTE]
这里的ndf并非指SQL文件中的ndf文件,它们不是同一个东西。
特性
支持语法高亮
支持语法错误提示
支持悬停提示
本地化语言支持
错误提示等级设置
支持全局符号构建和查找
可以解析文件中通过全局(未显式)导入的符号
支持自动注入全局符号导入(自动显式导入)
将已解析的全局符号的导入注释插入到文件的头部。即使不导入，全局符号查找功能也能识别全局符号，
导入只是为了加快解析速度，但在优化后只是锦上添花。
[!TIP]
导入注释语法:
- 从文件中导入:
\和/皆可
- 绝对导入:
from 'path\to\file.ndf' import symbol1, symbol2, ...
- 相对根目录导入:
from '\path\to\file.ndf' import symbol1, symbol2, ...
- 相对当前文件导入:
from '.\path\to\file.ndf' import symbol1, symbol2, ...
- 从
内置库导入:
目前测试的内置库符号只有TBaseClass,且暂不支持自动注入
支持语义高亮
支持代码自动补全
依赖
暂时没有依赖,如果非要说有的话,你可能需要一款使用该语言的游戏
扩展设置
设置位置: 文件 > 首选项 > 设置 > 用户 > 扩展 > NDF
language:
语言设置,默认为'en-US',目前仅支持'en-US'和'zh-CN'。一般切换语言后即刻生效,在全局构建任务正在运行的情况下可能会有延迟。
level:
错误提示等级设置,默认为'strict',可选'strict', 'loose', 'ignore'。一般切换等级后即刻生效,在全局构建任务正在运行的情况下可能会有延迟。切换后错误提示将会降级或升级。
maxNumberOfProblems:
最大错误提示数量,默认为100。
globalSearch:
全局符号搜索,默认为true。开启后将会在当前项目根目录下构建、缓存和查找全局符号。
当构建任务运行时,根目录下会生成.ndf-cache文件夹,里面包含了全局符号的缓存。
[!NOTE]
在构建期间,.ndf-cache文件夹中文件频繁创建和删除,这是正常现象
processNumber:
构建任务进程数,默认为3。
[!WARNING]
该设置可能会影响构建速度,但过多可能会导致不可预期的现象.
threadNumber:
每一个进程的线程数,默认为当前CPU核数/2。
[!WARNING]
该设置可能会影响构建速度,但过多可能会导致不可预期的现象.
asyncWorkerNumber:
每个线程的异步工作者数,默认为10。
autoSpawnImport:
自动注入全局符号导入,默认为true。开启后将会在文件头部注入全局符号导入注释。
已知问题
暂时没有已知问题,但如果有的话,请在issues中留言。
- 在文件过大时，解析时长过于漫长，导致打开其它文件时不会进行解析。
- 由于全局符号是在解析时是在文件作用域内进行的，所以在其定义中包含全局符号时，将会丧失关联信息。
待办事项
解析器
- [ ] 支持特殊构造函数的类型验证
- [ ] 编写内置库
IDE
- [ ] 添加自动补全功能
- [ ] 添加代码格式化功能
发布说明
这是版本3.0.0的发布说明。该版本随NDF语言新语法进行了适应性、破坏性更新。
由于Eugen System没有公布NDF的完整语法和语义，解析器目前的语法仅根据《ndf_reference_manual.pdf》与官方ndf语料逆推。
因此该分析器只能无限的接近真正的NDF解释/编译器,但不等于它。
由于性能问题基于C++的重制仓库目前处于private状态，因此原仓库不再支持。
NDF language is a scripting language used to describe game data.
[!NOTE]
The ndf here does not refer to the ndf file in the SQL file, they are not the same thing.
Features
Support for syntax highlighting
Support for syntax error reporting
Support for hover tooltips
Localized Language Support
Error Message Severity Level Setting
Support for global symbol building and lookup
Can parse symbols globally imported (implicitly) in the file
Support for automatic injection of global symbol imports (automatic explicit import)
Inserts parsed global symbol import comments into the file header. This means that even if the import is not
performed, the global symbol lookup function can recognize global symbols, and the import is only for speed optimization,
but it is a nice touch after optimization.
[!TIP]
Import comment syntax:
- Import from a file:
Both \ and / are acceptable
- Absolute import:
from 'path\to\file.ndf' import symbol1, symbol2, ...
- Relative to root directory import:
from '\path\to\file.ndf' import symbol1, symbol2, ...
- Relative to current file import:
from '.\path\to\file.ndf' import symbol1, symbol2, ...
- Import from
built-in library:
Currently, the only tested built-in library symbol is TBaseClass, and automatic injection is not supported.
Support for semantic highlighting
Support for code auto-completion
Requirements
There is currently no requirement, but if necessary, you may need a game that uses that language
Extension Settings
Location: File > Preferences > Settings > Extensions > NDF
language:
Language setting, default is 'en-US', currently supports 'en-US' and 'zh-CN'. Language changes usually take
effect immediately, but there may be a delay if a global build task is running.
level:
Error message level setting, default is 'strict', options include 'strict', 'loose', 'ignore'. Level changes
usually take effect immediately, but there may be a delay if a global build task is running. Changing the level will
downgrade or upgrade error messages accordingly.
maxNumberOfProblems:
Maximum number of error messages, default is 100.
globalSearch:
Global symbol search, default is true. Enabling this will build, cache, and search for global symbols in the root
directory of the current project.
When the build task is running, a .ndf-cache folder will be generated in the root directory, containing the cache of
global symbols.
[!NOTE]
During the build process, files in the .ndf-cache folder are frequently created and deleted, which is normal behavior.
processNumber:
Number of build task processes, default is 3.
[!WARNING]
This setting may affect build speed, but too many processes may lead to unpredictable behavior.
threadNumber:
Number of threads per process, default is half the number of CPU cores.
[!WARNING]
This setting may affect build speed, but too many threads may lead to unpredictable behavior.
asyncWorkerNumber:
Number of asynchronous workers per thread, default is 10.
autoSpawnImport:
Automatic injection of global symbol imports, default is false. Enabling this will inject global symbol import
comments into the file header.
Known Issues
please leave a comment in the issues section.
- When a file is too large, the parsing time is too long, which causes the opening of other files to not perform parsing.
- Because global symbols are defined in file scope, they will lose the association information when they contain global
symbols.
Todo
Parser
- [ ] Support type validation for special constructor functions
- [ ] Write built-in library
IDE
- [ ] Add code formatting functionality
Release Notes
These are the release notes for version 3.0.0. This version includes adaptive and breaking changes for the new NDF language syntax.
Since Eugen System has not published the complete syntax and semantics of NDF, the parser's current grammar is only reverse-engineered based on ndf_reference_manual.pdf and official NDF corpora.
Therefore, this analyzer can only infinitely approach the real NDF interpreter/compiler, but it is not equivalent to it.
Due to performance issues, the C++-based remake repository is currently private, so the original repository is no longer supported.
| |
