https://github.com/berquerant/dql
Find files or directories by sql.
https://github.com/berquerant/dql
go yacc
Last synced: about 1 year ago
JSON representation
Find files or directories by sql.
- Host: GitHub
- URL: https://github.com/berquerant/dql
- Owner: berquerant
- License: mit
- Created: 2021-10-12T17:28:33.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2022-07-18T06:15:27.000Z (almost 4 years ago)
- Last Synced: 2025-02-06T16:58:40.697Z (over 1 year ago)
- Topics: go, yacc
- Language: Go
- Homepage:
- Size: 70.3 KB
- Stars: 0
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# dql
Find files or directories by sql.
## Syntax
```
SELECT select_expr [, select_expr ...]
[WHERE where_condition]
[GROUP BY col_name]
[HAVING having_condition]
[ORDER BY order_by_expr [ASC | DESC]]
[LIMIT row_count [OFFSET offset]]
```
### SELECT
`select_expr` is a column that you want.
```
select name, size;
```
Select all columns.
```
select all;
```
is equivalent to
```
select name, size, mode, mod_time, is_dir;
```
Give a temporary name:
```
select name as N;
```
The temporary name is available in `WHERE`, `HAVING` and `ORDER BY`.
Aggregations are available with conditions below:
a. without `GROUP BY` and select aggregations only.
b. with `GROUP BY` then except `GROUP BY` column.
### WHERE
`where_condition` is a condition expr, if the evaluated value of a row is true then the row is selected.
```
select all where is_dir;
```
### GROUP BY
`GROUP BY` aggregates rows by `col_name`.
`SELECT` must contain `col_name` and not contain raw columns but `col_name`.
```
select is_dir, count(name) group by is_dir;
```
### HAVING
`having_condition` is a condition expr, if the evaluated value of a row is true then the row is selected.
`HAVING` must be written with `GROUP BY`.
```
select mode, count(name) group by mode having count(name) > 5;
```
### ORDER BY
`order_by_expr` is a expr on which to sort rows.
```
select len(name) as nlen, name order by nlen desc;
```
### LIMIT
`row_count` constrains the number of the result rows.
```
select all limit 3;
```
If `offset` is exist, ignore first `offset` rows.
```
select all limit 3 offset 5;
```
## Columns
- `name` is the path.
- `size` is the size, number of bytes in the file.
- `mode` is the file mode, the entry type and permissions.
- `mod_time` is the last modification time.
- `is_dir` is the limited entry type, if true, the row comes from a file.
## Data types
| Name | Description | Example |
|--------|----------------|---------------|
| int | integer | 10, -1 |
| float | floating point | 1.2, -0.5 |
| string | string | "str" |
| bool | bool | (no literals) |
Hereafter, int or float are referred to as number,
and a string literal matched with `[01]+` is referred to as bits.
## Operators
The operators in the more lower row has the higher precedence.
| Format | Description | Argument Types | Result Type | Example |
|-----------------------|-----------------------|---------------------------|-------------|-----------------------------|
| or | or | bool, bool | bool | is_dir or size < 100 |
| and | and | bool, bool | bool | is_dir and name = "x" |
| xor | xor | bool, bool | bool | size < 100 xor name = "x" |
| = | equal | any, any (same type) | bool | name = "x" |
| <> | not equal | any, any (same type) | bool | name <> "x" |
| < | less than | any, any (same type) | bool | size < 100 |
| <= | less than or equal | any, any (same type) | bool | size <= 100 |
| > | greater than | any, any (same type) | bool | size > 100 |
| >= | greater than or equal | any, any (same type) | bool | size >= 100 |
| . in (...) | within | any, list of any | bool | name in ("x", "y") |
| . not in (...) | not within | any, list of any | bool | name not in ("x", "y") |
| . between . and . | between | any, any, any (same type) | bool | size between 10 and 100 |
| . not between . and . | not between | any, any, any (same type) | bool | size not between 10 and 100 |
| . like REGEX | matched string | string, string | bool | name like "log$" |
| . not like REGEX | not matched string | string, string | bool | name not like "log$" |
| + | add | number, number | number | size + 1 |
| \- | subtract | number, number | number | size - 1 |
| \* | multiply | number, number | number | size * 2 |
| / | division | number, number | number | size / 2 |
| \| | bit or | int or bits, int or bits | int | 3 \| 4 |
| & | bit and | int or bits, int or bits | int | 3 & 4 |
| ^ | bit xor | int or bits, int or bits | int | 3 ^ 4 |
| + | noop | any | any | +1 |
| \- | unary minus | number | number | -1 |
| ~ | bit not | int or bits | int | ~15 |
| not | not | bool | bool | not is_dir |
## Functions
The function converts an expr or a column into some value.
| Format | Description | Argument Types | Result Type | Example |
|------------|----------------------------------|----------------|-------------|--------------------------|
| pow(x, y) | x to the power of y | number, number | number | pow(2, 3) |
| ceil(x) | ceiling | number | int | ceil(2.3) |
| floor(x) | floor | number | int | floor(2.3) |
| len(x) | length of string | string | int | len("length") |
| base(x) | the last element of path | string | string | base("dir/file") |
| dir(x) | all but the last element of path | string | string | dir("dir/file") |
| ext(x) | the file name extension | string | string | ext("dired.elc") |
| bin2int(x) | bits to int | bits | int | bin2int("1010") |
| int2bin(x) | int to bits | int | bits | int2bin(10) |
| cast(x, y) | cast x to y | any | string | cast(10, "string") |
| now() | the current local time | | int | now() |
| depth(x) | the depth of the path | name | int | depth("/home/user") |
| grep(x, y) | `grep x y` | string | string | grep("lambda", "map.py") |
### Cast
`cast(value, "destination type")` cast value to destination type.
If the type of the value equals the destination type, then the result is the value itself.
More conversion rules are below:
| Value Type | Destination Type | Description |
|------------|------------------|------------------------------------------------------------------------------|
| float | int | floor |
| string | int | parse |
| bool | int | true into 1, false into 0 |
| int | float | type conversion only |
| string | float | parse |
| bool | float | true into 1.0, false into 0.0 |
| any | string | format as string |
| number | bool | value != 0 |
| string | bool | value != "" |
| string | timestamp | parse string like "2006-01-02T15:04:05Z07:00" (RFC3339) into timestamp (int) |
| number | time | timestamp into time (string) |
| number | duration | seconds into duration (string) |
Undefined rules cause conversion errors.
### Aggregations
The aggregation converts the rows into some value.
| Format | Description | Argument Types | Result Type | Example |
|------------|-----------------------|----------------|-----------------|---------------|
| count(x) | number of the rows | any | int | count(name) |
| min(x) | minimum of the rows | any | any (same type) | min(size) |
| max(x) | maximum of the rows | any | any (same type) | max(name) |
| product(x) | product of the rows | number | number | product(size) |
| sum(x) | summation of the rows | number | number | sum(size) |
| avg(x) | average of the rows | number | number | avg(size) |
## Reserved words
The reserved words are case insensitive.
```
select where having group by order limit as asc desc like in not and or xor between offset
```
## Usage
```
make dist/dql
cd dist
dql -h
```