1
0
Fork 0
mirror of https://github.com/moby/moby.git synced 2022-11-09 12:21:53 -05:00
moby--moby/builder/dockerfile/parser/parser.go

222 lines
6.8 KiB
Go
Raw Normal View History

// Package parser implements a parser and parse tree dumper for Dockerfiles.
package parser
import (
"bufio"
"bytes"
"fmt"
"io"
"regexp"
"strings"
"unicode"
"github.com/docker/docker/builder/dockerfile/command"
)
// Node is a structure used to represent a parse tree.
//
// In the node there are three fields, Value, Next, and Children. Value is the
// current token's string value. Next is always the next non-child token, and
// children contains all the children. Here's an example:
//
// (value next (child child-next child-next-next) next-next)
//
// This data structure is frankly pretty lousy for handling complex languages,
// but lucky for us the Dockerfile isn't very complicated. This structure
// works a little more effectively than a "proper" parse tree for our needs.
//
type Node struct {
Value string // actual content
Next *Node // the next item in the current sexp
Children []*Node // the children of this sexp
Attributes map[string]bool // special attributes for this node
Original string // original line used before parsing
Flags []string // only top Node should have this set
StartLine int // the line in the original dockerfile where the node begins
EndLine int // the line in the original dockerfile where the node ends
}
// Directive is the structure used during a build run to hold the state of
// parsing directives.
type Directive struct {
EscapeToken rune // Current escape token
LineContinuationRegex *regexp.Regexp // Current line continuation regex
LookingForDirectives bool // Whether we are currently looking for directives
EscapeSeen bool // Whether the escape directive has been seen
}
var (
dispatch map[string]func(string, *Directive) (*Node, map[string]bool, error)
tokenWhitespace = regexp.MustCompile(`[\t\v\f\r ]+`)
tokenEscapeCommand = regexp.MustCompile(`^#[ \t]*escape[ \t]*=[ \t]*(?P<escapechar>.).*$`)
tokenComment = regexp.MustCompile(`^#.*$`)
)
// DefaultEscapeToken is the default escape token
const DefaultEscapeToken = "\\"
// SetEscapeToken sets the default token for escaping characters in a Dockerfile.
func SetEscapeToken(s string, d *Directive) error {
if s != "`" && s != "\\" {
return fmt.Errorf("invalid ESCAPE '%s'. Must be ` or \\", s)
}
d.EscapeToken = rune(s[0])
d.LineContinuationRegex = regexp.MustCompile(`\` + s + `[ \t]*$`)
return nil
}
func init() {
// Dispatch Table. see line_parsers.go for the parse functions.
// The command is parsed and mapped to the line parser. The line parser
// receives the arguments but not the command, and returns an AST after
// reformulating the arguments according to the rules in the parser
// functions. Errors are propagated up by Parse() and the resulting AST can
// be incorporated directly into the existing AST as a next.
dispatch = map[string]func(string, *Directive) (*Node, map[string]bool, error){
Add support for user-defined healthchecks This PR adds support for user-defined health-check probes for Docker containers. It adds a `HEALTHCHECK` instruction to the Dockerfile syntax plus some corresponding "docker run" options. It can be used with a restart policy to automatically restart a container if the check fails. The `HEALTHCHECK` instruction has two forms: * `HEALTHCHECK [OPTIONS] CMD command` (check container health by running a command inside the container) * `HEALTHCHECK NONE` (disable any healthcheck inherited from the base image) The `HEALTHCHECK` instruction tells Docker how to test a container to check that it is still working. This can detect cases such as a web server that is stuck in an infinite loop and unable to handle new connections, even though the server process is still running. When a container has a healthcheck specified, it has a _health status_ in addition to its normal status. This status is initially `starting`. Whenever a health check passes, it becomes `healthy` (whatever state it was previously in). After a certain number of consecutive failures, it becomes `unhealthy`. The options that can appear before `CMD` are: * `--interval=DURATION` (default: `30s`) * `--timeout=DURATION` (default: `30s`) * `--retries=N` (default: `1`) The health check will first run **interval** seconds after the container is started, and then again **interval** seconds after each previous check completes. If a single run of the check takes longer than **timeout** seconds then the check is considered to have failed. It takes **retries** consecutive failures of the health check for the container to be considered `unhealthy`. There can only be one `HEALTHCHECK` instruction in a Dockerfile. If you list more than one then only the last `HEALTHCHECK` will take effect. The command after the `CMD` keyword can be either a shell command (e.g. `HEALTHCHECK CMD /bin/check-running`) or an _exec_ array (as with other Dockerfile commands; see e.g. `ENTRYPOINT` for details). The command's exit status indicates the health status of the container. The possible values are: - 0: success - the container is healthy and ready for use - 1: unhealthy - the container is not working correctly - 2: starting - the container is not ready for use yet, but is working correctly If the probe returns 2 ("starting") when the container has already moved out of the "starting" state then it is treated as "unhealthy" instead. For example, to check every five minutes or so that a web-server is able to serve the site's main page within three seconds: HEALTHCHECK --interval=5m --timeout=3s \ CMD curl -f http://localhost/ || exit 1 To help debug failing probes, any output text (UTF-8 encoded) that the command writes on stdout or stderr will be stored in the health status and can be queried with `docker inspect`. Such output should be kept short (only the first 4096 bytes are stored currently). When the health status of a container changes, a `health_status` event is generated with the new status. The health status is also displayed in the `docker ps` output. Signed-off-by: Thomas Leonard <thomas.leonard@docker.com> Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2016-04-18 05:48:13 -04:00
command.Add: parseMaybeJSONToList,
command.Arg: parseNameOrNameVal,
Add support for user-defined healthchecks This PR adds support for user-defined health-check probes for Docker containers. It adds a `HEALTHCHECK` instruction to the Dockerfile syntax plus some corresponding "docker run" options. It can be used with a restart policy to automatically restart a container if the check fails. The `HEALTHCHECK` instruction has two forms: * `HEALTHCHECK [OPTIONS] CMD command` (check container health by running a command inside the container) * `HEALTHCHECK NONE` (disable any healthcheck inherited from the base image) The `HEALTHCHECK` instruction tells Docker how to test a container to check that it is still working. This can detect cases such as a web server that is stuck in an infinite loop and unable to handle new connections, even though the server process is still running. When a container has a healthcheck specified, it has a _health status_ in addition to its normal status. This status is initially `starting`. Whenever a health check passes, it becomes `healthy` (whatever state it was previously in). After a certain number of consecutive failures, it becomes `unhealthy`. The options that can appear before `CMD` are: * `--interval=DURATION` (default: `30s`) * `--timeout=DURATION` (default: `30s`) * `--retries=N` (default: `1`) The health check will first run **interval** seconds after the container is started, and then again **interval** seconds after each previous check completes. If a single run of the check takes longer than **timeout** seconds then the check is considered to have failed. It takes **retries** consecutive failures of the health check for the container to be considered `unhealthy`. There can only be one `HEALTHCHECK` instruction in a Dockerfile. If you list more than one then only the last `HEALTHCHECK` will take effect. The command after the `CMD` keyword can be either a shell command (e.g. `HEALTHCHECK CMD /bin/check-running`) or an _exec_ array (as with other Dockerfile commands; see e.g. `ENTRYPOINT` for details). The command's exit status indicates the health status of the container. The possible values are: - 0: success - the container is healthy and ready for use - 1: unhealthy - the container is not working correctly - 2: starting - the container is not ready for use yet, but is working correctly If the probe returns 2 ("starting") when the container has already moved out of the "starting" state then it is treated as "unhealthy" instead. For example, to check every five minutes or so that a web-server is able to serve the site's main page within three seconds: HEALTHCHECK --interval=5m --timeout=3s \ CMD curl -f http://localhost/ || exit 1 To help debug failing probes, any output text (UTF-8 encoded) that the command writes on stdout or stderr will be stored in the health status and can be queried with `docker inspect`. Such output should be kept short (only the first 4096 bytes are stored currently). When the health status of a container changes, a `health_status` event is generated with the new status. The health status is also displayed in the `docker ps` output. Signed-off-by: Thomas Leonard <thomas.leonard@docker.com> Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2016-04-18 05:48:13 -04:00
command.Cmd: parseMaybeJSON,
command.Copy: parseMaybeJSONToList,
Add support for user-defined healthchecks This PR adds support for user-defined health-check probes for Docker containers. It adds a `HEALTHCHECK` instruction to the Dockerfile syntax plus some corresponding "docker run" options. It can be used with a restart policy to automatically restart a container if the check fails. The `HEALTHCHECK` instruction has two forms: * `HEALTHCHECK [OPTIONS] CMD command` (check container health by running a command inside the container) * `HEALTHCHECK NONE` (disable any healthcheck inherited from the base image) The `HEALTHCHECK` instruction tells Docker how to test a container to check that it is still working. This can detect cases such as a web server that is stuck in an infinite loop and unable to handle new connections, even though the server process is still running. When a container has a healthcheck specified, it has a _health status_ in addition to its normal status. This status is initially `starting`. Whenever a health check passes, it becomes `healthy` (whatever state it was previously in). After a certain number of consecutive failures, it becomes `unhealthy`. The options that can appear before `CMD` are: * `--interval=DURATION` (default: `30s`) * `--timeout=DURATION` (default: `30s`) * `--retries=N` (default: `1`) The health check will first run **interval** seconds after the container is started, and then again **interval** seconds after each previous check completes. If a single run of the check takes longer than **timeout** seconds then the check is considered to have failed. It takes **retries** consecutive failures of the health check for the container to be considered `unhealthy`. There can only be one `HEALTHCHECK` instruction in a Dockerfile. If you list more than one then only the last `HEALTHCHECK` will take effect. The command after the `CMD` keyword can be either a shell command (e.g. `HEALTHCHECK CMD /bin/check-running`) or an _exec_ array (as with other Dockerfile commands; see e.g. `ENTRYPOINT` for details). The command's exit status indicates the health status of the container. The possible values are: - 0: success - the container is healthy and ready for use - 1: unhealthy - the container is not working correctly - 2: starting - the container is not ready for use yet, but is working correctly If the probe returns 2 ("starting") when the container has already moved out of the "starting" state then it is treated as "unhealthy" instead. For example, to check every five minutes or so that a web-server is able to serve the site's main page within three seconds: HEALTHCHECK --interval=5m --timeout=3s \ CMD curl -f http://localhost/ || exit 1 To help debug failing probes, any output text (UTF-8 encoded) that the command writes on stdout or stderr will be stored in the health status and can be queried with `docker inspect`. Such output should be kept short (only the first 4096 bytes are stored currently). When the health status of a container changes, a `health_status` event is generated with the new status. The health status is also displayed in the `docker ps` output. Signed-off-by: Thomas Leonard <thomas.leonard@docker.com> Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2016-04-18 05:48:13 -04:00
command.Entrypoint: parseMaybeJSON,
command.Env: parseEnv,
Add support for user-defined healthchecks This PR adds support for user-defined health-check probes for Docker containers. It adds a `HEALTHCHECK` instruction to the Dockerfile syntax plus some corresponding "docker run" options. It can be used with a restart policy to automatically restart a container if the check fails. The `HEALTHCHECK` instruction has two forms: * `HEALTHCHECK [OPTIONS] CMD command` (check container health by running a command inside the container) * `HEALTHCHECK NONE` (disable any healthcheck inherited from the base image) The `HEALTHCHECK` instruction tells Docker how to test a container to check that it is still working. This can detect cases such as a web server that is stuck in an infinite loop and unable to handle new connections, even though the server process is still running. When a container has a healthcheck specified, it has a _health status_ in addition to its normal status. This status is initially `starting`. Whenever a health check passes, it becomes `healthy` (whatever state it was previously in). After a certain number of consecutive failures, it becomes `unhealthy`. The options that can appear before `CMD` are: * `--interval=DURATION` (default: `30s`) * `--timeout=DURATION` (default: `30s`) * `--retries=N` (default: `1`) The health check will first run **interval** seconds after the container is started, and then again **interval** seconds after each previous check completes. If a single run of the check takes longer than **timeout** seconds then the check is considered to have failed. It takes **retries** consecutive failures of the health check for the container to be considered `unhealthy`. There can only be one `HEALTHCHECK` instruction in a Dockerfile. If you list more than one then only the last `HEALTHCHECK` will take effect. The command after the `CMD` keyword can be either a shell command (e.g. `HEALTHCHECK CMD /bin/check-running`) or an _exec_ array (as with other Dockerfile commands; see e.g. `ENTRYPOINT` for details). The command's exit status indicates the health status of the container. The possible values are: - 0: success - the container is healthy and ready for use - 1: unhealthy - the container is not working correctly - 2: starting - the container is not ready for use yet, but is working correctly If the probe returns 2 ("starting") when the container has already moved out of the "starting" state then it is treated as "unhealthy" instead. For example, to check every five minutes or so that a web-server is able to serve the site's main page within three seconds: HEALTHCHECK --interval=5m --timeout=3s \ CMD curl -f http://localhost/ || exit 1 To help debug failing probes, any output text (UTF-8 encoded) that the command writes on stdout or stderr will be stored in the health status and can be queried with `docker inspect`. Such output should be kept short (only the first 4096 bytes are stored currently). When the health status of a container changes, a `health_status` event is generated with the new status. The health status is also displayed in the `docker ps` output. Signed-off-by: Thomas Leonard <thomas.leonard@docker.com> Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2016-04-18 05:48:13 -04:00
command.Expose: parseStringsWhitespaceDelimited,
command.From: parseString,
Add support for user-defined healthchecks This PR adds support for user-defined health-check probes for Docker containers. It adds a `HEALTHCHECK` instruction to the Dockerfile syntax plus some corresponding "docker run" options. It can be used with a restart policy to automatically restart a container if the check fails. The `HEALTHCHECK` instruction has two forms: * `HEALTHCHECK [OPTIONS] CMD command` (check container health by running a command inside the container) * `HEALTHCHECK NONE` (disable any healthcheck inherited from the base image) The `HEALTHCHECK` instruction tells Docker how to test a container to check that it is still working. This can detect cases such as a web server that is stuck in an infinite loop and unable to handle new connections, even though the server process is still running. When a container has a healthcheck specified, it has a _health status_ in addition to its normal status. This status is initially `starting`. Whenever a health check passes, it becomes `healthy` (whatever state it was previously in). After a certain number of consecutive failures, it becomes `unhealthy`. The options that can appear before `CMD` are: * `--interval=DURATION` (default: `30s`) * `--timeout=DURATION` (default: `30s`) * `--retries=N` (default: `1`) The health check will first run **interval** seconds after the container is started, and then again **interval** seconds after each previous check completes. If a single run of the check takes longer than **timeout** seconds then the check is considered to have failed. It takes **retries** consecutive failures of the health check for the container to be considered `unhealthy`. There can only be one `HEALTHCHECK` instruction in a Dockerfile. If you list more than one then only the last `HEALTHCHECK` will take effect. The command after the `CMD` keyword can be either a shell command (e.g. `HEALTHCHECK CMD /bin/check-running`) or an _exec_ array (as with other Dockerfile commands; see e.g. `ENTRYPOINT` for details). The command's exit status indicates the health status of the container. The possible values are: - 0: success - the container is healthy and ready for use - 1: unhealthy - the container is not working correctly - 2: starting - the container is not ready for use yet, but is working correctly If the probe returns 2 ("starting") when the container has already moved out of the "starting" state then it is treated as "unhealthy" instead. For example, to check every five minutes or so that a web-server is able to serve the site's main page within three seconds: HEALTHCHECK --interval=5m --timeout=3s \ CMD curl -f http://localhost/ || exit 1 To help debug failing probes, any output text (UTF-8 encoded) that the command writes on stdout or stderr will be stored in the health status and can be queried with `docker inspect`. Such output should be kept short (only the first 4096 bytes are stored currently). When the health status of a container changes, a `health_status` event is generated with the new status. The health status is also displayed in the `docker ps` output. Signed-off-by: Thomas Leonard <thomas.leonard@docker.com> Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
2016-04-18 05:48:13 -04:00
command.Healthcheck: parseHealthConfig,
command.Label: parseLabel,
command.Maintainer: parseString,
command.Onbuild: parseSubCommand,
command.Run: parseMaybeJSON,
command.Shell: parseMaybeJSON,
command.StopSignal: parseString,
command.User: parseString,
command.Volume: parseMaybeJSONToList,
command.Workdir: parseString,
}
}
// ParseLine parses a line and returns the remainder.
func ParseLine(line string, d *Directive, ignoreCont bool) (string, *Node, error) {
// Handle the parser directive '# escape=<char>. Parser directives must precede
// any builder instruction or other comments, and cannot be repeated.
if d.LookingForDirectives {
tecMatch := tokenEscapeCommand.FindStringSubmatch(strings.ToLower(line))
if len(tecMatch) > 0 {
if d.EscapeSeen == true {
return "", nil, fmt.Errorf("only one escape parser directive can be used")
}
for i, n := range tokenEscapeCommand.SubexpNames() {
if n == "escapechar" {
if err := SetEscapeToken(tecMatch[i], d); err != nil {
return "", nil, err
}
d.EscapeSeen = true
return "", nil, nil
}
}
}
}
d.LookingForDirectives = false
if line = stripComments(line); line == "" {
return "", nil, nil
}
if !ignoreCont && d.LineContinuationRegex.MatchString(line) {
line = d.LineContinuationRegex.ReplaceAllString(line, "")
return line, nil, nil
}
cmd, flags, args, err := splitCommand(line)
if err != nil {
return "", nil, err
}
node := &Node{}
node.Value = cmd
sexp, attrs, err := fullDispatch(cmd, args, d)
if err != nil {
return "", nil, err
}
node.Next = sexp
node.Attributes = attrs
node.Original = line
node.Flags = flags
return "", node, nil
}
// Parse is the main parse routine.
// It handles an io.ReadWriteCloser and returns the root of the AST.
func Parse(rwc io.Reader, d *Directive) (*Node, error) {
currentLine := 0
root := &Node{}
root.StartLine = -1
scanner := bufio.NewScanner(rwc)
utf8bom := []byte{0xEF, 0xBB, 0xBF}
for scanner.Scan() {
scannedBytes := scanner.Bytes()
// We trim UTF8 BOM
if currentLine == 0 {
scannedBytes = bytes.TrimPrefix(scannedBytes, utf8bom)
}
scannedLine := strings.TrimLeftFunc(string(scannedBytes), unicode.IsSpace)
currentLine++
line, child, err := ParseLine(scannedLine, d, false)
if err != nil {
return nil, err
}
startLine := currentLine
if line != "" && child == nil {
for scanner.Scan() {
newline := scanner.Text()
currentLine++
if stripComments(strings.TrimSpace(newline)) == "" {
continue
}
line, child, err = ParseLine(line+newline, d, false)
if err != nil {
return nil, err
}
if child != nil {
break
}
}
if child == nil && line != "" {
// When we call ParseLine we'll pass in 'true' for
// the ignoreCont param if we're at the EOF. This will
// prevent the func from returning immediately w/o
// parsing the line thinking that there's more input
// to come.
_, child, err = ParseLine(line, d, scanner.Err() == nil)
if err != nil {
return nil, err
}
}
}
if child != nil {
// Update the line information for the current child.
child.StartLine = startLine
child.EndLine = currentLine
// Update the line information for the root. The starting line of the root is always the
// starting line of the first child and the ending line is the ending line of the last child.
if root.StartLine < 0 {
root.StartLine = currentLine
}
root.EndLine = currentLine
root.Children = append(root.Children, child)
}
}
return root, nil
}