forked from lobo/lambda-zig
refactor main loop into multiple functions
This also gains the ability to report errors when posting the response
This commit is contained in:
parent
883bc6f52f
commit
e2bbaf2c9a
395
src/lambda.zig
395
src/lambda.zig
|
@ -5,11 +5,21 @@ const HandlerFn = *const fn (std.mem.Allocator, []const u8) anyerror![]const u8;
|
||||||
|
|
||||||
const log = std.log.scoped(.lambda);
|
const log = std.log.scoped(.lambda);
|
||||||
|
|
||||||
|
var empty_headers: std.http.Headers = undefined;
|
||||||
|
var client: ?std.http.Client = null;
|
||||||
|
|
||||||
|
const prefix = "http://";
|
||||||
|
const postfix = "/2018-06-01/runtime/invocation";
|
||||||
|
|
||||||
|
pub fn deinit() void {
|
||||||
|
if (client) |*c| c.deinit();
|
||||||
|
client = null;
|
||||||
|
}
|
||||||
/// Starts the lambda framework. Handler will be called when an event is processing
|
/// Starts the lambda framework. Handler will be called when an event is processing
|
||||||
/// If an allocator is not provided, an approrpriate allocator will be selected and used
|
/// If an allocator is not provided, an approrpriate allocator will be selected and used
|
||||||
|
/// This function is intended to loop infinitely. If not used in this manner,
|
||||||
|
/// make sure to call the deinit() function
|
||||||
pub fn run(allocator: ?std.mem.Allocator, event_handler: HandlerFn) !void { // TODO: remove inferred error set?
|
pub fn run(allocator: ?std.mem.Allocator, event_handler: HandlerFn) !void { // TODO: remove inferred error set?
|
||||||
const prefix = "http://";
|
|
||||||
const postfix = "/2018-06-01/runtime/invocation";
|
|
||||||
const lambda_runtime_uri = std.os.getenv("AWS_LAMBDA_RUNTIME_API") orelse test_lambda_runtime_uri.?;
|
const lambda_runtime_uri = std.os.getenv("AWS_LAMBDA_RUNTIME_API") orelse test_lambda_runtime_uri.?;
|
||||||
// TODO: If this is null, go into single use command line mode
|
// TODO: If this is null, go into single use command line mode
|
||||||
|
|
||||||
|
@ -21,9 +31,13 @@ pub fn run(allocator: ?std.mem.Allocator, event_handler: HandlerFn) !void { // T
|
||||||
defer alloc.free(url);
|
defer alloc.free(url);
|
||||||
const uri = try std.Uri.parse(url);
|
const uri = try std.Uri.parse(url);
|
||||||
|
|
||||||
var client: std.http.Client = .{ .allocator = alloc };
|
// TODO: Simply adding this line without even using the client is enough
|
||||||
defer client.deinit();
|
// to cause seg faults!?
|
||||||
var empty_headers = std.http.Headers.init(alloc);
|
// client = client orelse .{ .allocator = alloc };
|
||||||
|
// so we'll do this instead
|
||||||
|
if (client != null) return error.MustDeInitBeforeCallingRunAgain;
|
||||||
|
client = .{ .allocator = alloc };
|
||||||
|
empty_headers = std.http.Headers.init(alloc);
|
||||||
defer empty_headers.deinit();
|
defer empty_headers.deinit();
|
||||||
log.info("tid {d} (lambda): Bootstrap initializing with event url: {s}", .{ std.Thread.getCurrentId(), url });
|
log.info("tid {d} (lambda): Bootstrap initializing with event url: {s}", .{ std.Thread.getCurrentId(), url });
|
||||||
|
|
||||||
|
@ -36,160 +50,229 @@ pub fn run(allocator: ?std.mem.Allocator, event_handler: HandlerFn) !void { // T
|
||||||
var req_alloc = std.heap.ArenaAllocator.init(alloc);
|
var req_alloc = std.heap.ArenaAllocator.init(alloc);
|
||||||
defer req_alloc.deinit();
|
defer req_alloc.deinit();
|
||||||
const req_allocator = req_alloc.allocator();
|
const req_allocator = req_alloc.allocator();
|
||||||
var req = try client.request(.GET, uri, empty_headers, .{});
|
|
||||||
defer req.deinit();
|
|
||||||
|
|
||||||
req.start() catch |err| { // Well, at this point all we can do is shout at the void
|
// Fundamentally we're doing 3 things:
|
||||||
log.err("Get fail (start): {}", .{err});
|
// 1. Get the next event from Lambda (event data and request id)
|
||||||
std.os.exit(0);
|
// 2. Call our handler to get the response
|
||||||
|
// 3. Post the response back to Lambda
|
||||||
|
var ev = getEvent(req_allocator, uri) catch |err| {
|
||||||
|
// Well, at this point all we can do is shout at the void
|
||||||
|
log.err("Error fetching event details: {}", .{err});
|
||||||
|
std.os.exit(1);
|
||||||
|
// continue;
|
||||||
|
};
|
||||||
|
if (ev == null) continue; // this gets logged in getEvent, and without
|
||||||
|
// a request id, we still can't do anything
|
||||||
|
// reasonable to report back
|
||||||
|
const event = ev.?;
|
||||||
|
defer ev.?.deinit();
|
||||||
|
const event_response = event_handler(req_allocator, event.event_data) catch |err| {
|
||||||
|
event.reportError(@errorReturnTrace(), err, lambda_runtime_uri) catch unreachable;
|
||||||
continue;
|
continue;
|
||||||
};
|
};
|
||||||
|
event.postResponse(lambda_runtime_uri, event_response) catch |err| {
|
||||||
// Lambda freezes the process at this line of code. During warm start,
|
event.reportError(@errorReturnTrace(), err, lambda_runtime_uri) catch unreachable;
|
||||||
// the process will unfreeze and data will be sent in response to client.get
|
|
||||||
req.wait() catch |err| { // Well, at this point all we can do is shout at the void
|
|
||||||
log.err("Get fail (wait): {}", .{err});
|
|
||||||
std.os.exit(0);
|
|
||||||
continue;
|
continue;
|
||||||
};
|
};
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
const Event = struct {
|
||||||
|
allocator: std.mem.Allocator,
|
||||||
|
event_data: []u8,
|
||||||
|
request_id: []u8,
|
||||||
|
|
||||||
|
const Self = @This();
|
||||||
|
|
||||||
|
pub fn init(allocator: std.mem.Allocator, event_data: []u8, request_id: []u8) Self {
|
||||||
|
return .{
|
||||||
|
.allocator = allocator,
|
||||||
|
.event_data = event_data,
|
||||||
|
.request_id = request_id,
|
||||||
|
};
|
||||||
|
}
|
||||||
|
pub fn deinit(self: *Self) void {
|
||||||
|
self.allocator.free(self.event_data);
|
||||||
|
self.allocator.free(self.request_id);
|
||||||
|
}
|
||||||
|
fn reportError(
|
||||||
|
self: Self,
|
||||||
|
return_trace: ?*std.builtin.StackTrace,
|
||||||
|
err: anytype,
|
||||||
|
lambda_runtime_uri: []const u8,
|
||||||
|
) !void {
|
||||||
|
// If we fail in this function, we're pretty hosed up
|
||||||
|
if (return_trace) |rt|
|
||||||
|
log.err("Caught error: {}. Return Trace: {any}", .{ err, rt })
|
||||||
|
else
|
||||||
|
log.err("Caught error: {}. No return trace available", .{err});
|
||||||
|
const err_url = try std.fmt.allocPrint(
|
||||||
|
self.allocator,
|
||||||
|
"{s}{s}{s}/{s}/error",
|
||||||
|
.{ prefix, lambda_runtime_uri, postfix, self.request_id },
|
||||||
|
);
|
||||||
|
defer self.allocator.free(err_url);
|
||||||
|
const err_uri = try std.Uri.parse(err_url);
|
||||||
|
const content =
|
||||||
|
\\{{
|
||||||
|
\\ "errorMessage": "{s}",
|
||||||
|
\\ "errorType": "HandlerReturnedError",
|
||||||
|
\\ "stackTrace": [ "{any}" ]
|
||||||
|
\\}}
|
||||||
|
;
|
||||||
|
const content_fmt = if (return_trace) |rt|
|
||||||
|
try std.fmt.allocPrint(self.allocator, content, .{ @errorName(err), rt })
|
||||||
|
else
|
||||||
|
try std.fmt.allocPrint(self.allocator, content, .{ @errorName(err), "no return trace available" });
|
||||||
|
defer self.allocator.free(content_fmt);
|
||||||
|
log.err("Posting to {s}: Data {s}", .{ err_url, content_fmt });
|
||||||
|
|
||||||
|
var err_headers = std.http.Headers.init(self.allocator);
|
||||||
|
defer err_headers.deinit();
|
||||||
|
err_headers.append(
|
||||||
|
"Lambda-Runtime-Function-Error-Type",
|
||||||
|
"HandlerReturned",
|
||||||
|
) catch |append_err| {
|
||||||
|
log.err("Error appending error header to post response for request id {s}: {}", .{ self.request_id, append_err });
|
||||||
|
std.os.exit(1);
|
||||||
|
};
|
||||||
|
// TODO: There is something up with using a shared client in this way
|
||||||
|
// so we're taking a perf hit in favor of stability. In a practical
|
||||||
|
// sense, without making HTTPS connections (lambda environment is
|
||||||
|
// non-ssl), this shouldn't be a big issue
|
||||||
|
var cl = std.http.Client{ .allocator = self.allocator };
|
||||||
|
defer cl.deinit();
|
||||||
|
var req = try cl.request(.POST, err_uri, empty_headers, .{});
|
||||||
|
// var req = try client.?.request(.POST, err_uri, empty_headers, .{});
|
||||||
|
// defer req.deinit();
|
||||||
|
req.transfer_encoding = .{ .content_length = content_fmt.len };
|
||||||
|
req.start() catch |post_err| { // Well, at this point all we can do is shout at the void
|
||||||
|
log.err("Error posting response (start) for request id {s}: {}", .{ self.request_id, post_err });
|
||||||
|
std.os.exit(1);
|
||||||
|
};
|
||||||
|
try req.writeAll(content_fmt);
|
||||||
|
try req.finish();
|
||||||
|
req.wait() catch |post_err| { // Well, at this point all we can do is shout at the void
|
||||||
|
log.err("Error posting response (wait) for request id {s}: {}", .{ self.request_id, post_err });
|
||||||
|
std.os.exit(1);
|
||||||
|
};
|
||||||
|
// TODO: Determine why this post is not returning
|
||||||
if (req.response.status != .ok) {
|
if (req.response.status != .ok) {
|
||||||
// Documentation says something about "exit immediately". The
|
// Documentation says something about "exit immediately". The
|
||||||
// Lambda infrastrucutre restarts, so it's unclear if that's necessary.
|
// Lambda infrastrucutre restarts, so it's unclear if that's necessary.
|
||||||
// It seems as though a continue should be fine, and slightly faster
|
// It seems as though a continue should be fine, and slightly faster
|
||||||
// std.os.exit(1);
|
|
||||||
log.err("Get fail: {} {s}", .{
|
log.err("Get fail: {} {s}", .{
|
||||||
@intFromEnum(req.response.status),
|
@intFromEnum(req.response.status),
|
||||||
req.response.status.phrase() orelse "",
|
req.response.status.phrase() orelse "",
|
||||||
});
|
});
|
||||||
continue;
|
std.os.exit(1);
|
||||||
}
|
}
|
||||||
|
log.err("Error reporting post complete", .{});
|
||||||
var request_id: ?[]const u8 = null;
|
|
||||||
var content_length: ?usize = null;
|
|
||||||
for (req.response.headers.list.items) |h| {
|
|
||||||
if (std.ascii.eqlIgnoreCase(h.name, "Lambda-Runtime-Aws-Request-Id"))
|
|
||||||
request_id = h.value;
|
|
||||||
if (std.ascii.eqlIgnoreCase(h.name, "Content-Length")) {
|
|
||||||
content_length = std.fmt.parseUnsigned(usize, h.value, 10) catch null;
|
|
||||||
if (content_length == null)
|
|
||||||
log.warn("Error parsing content length value: '{s}'", .{h.value});
|
|
||||||
}
|
|
||||||
// TODO: XRay uses an environment variable to do its magic. It's our
|
|
||||||
// responsibility to set this, but no zig-native setenv(3)/putenv(3)
|
|
||||||
// exists. I would kind of rather not link in libc for this,
|
|
||||||
// so we'll hold for now and think on this
|
|
||||||
// if (std.mem.indexOf(u8, h.name.value, "Lambda-Runtime-Trace-Id")) |_|
|
|
||||||
// std.process.
|
|
||||||
// std.os.setenv("AWS_LAMBDA_RUNTIME_API");
|
|
||||||
}
|
|
||||||
if (request_id == null) {
|
|
||||||
// We can't report back an issue because the runtime error reporting endpoint
|
|
||||||
// uses request id in its path. So the best we can do is log the error and move
|
|
||||||
// on here.
|
|
||||||
log.err("Could not find request id: skipping request", .{});
|
|
||||||
continue;
|
|
||||||
}
|
|
||||||
const req_id = request_id.?;
|
|
||||||
log.debug("got lambda request with id {s}", .{req_id});
|
|
||||||
|
|
||||||
const reader = req.reader();
|
|
||||||
var buf: [65535]u8 = undefined;
|
|
||||||
|
|
||||||
var resp_payload = std.ArrayList(u8).init(req_allocator);
|
|
||||||
if (content_length) |len| {
|
|
||||||
resp_payload.ensureTotalCapacity(len) catch {
|
|
||||||
log.err("Could not allocate memory for body of request id: {s}", .{request_id.?});
|
|
||||||
continue;
|
|
||||||
};
|
|
||||||
}
|
|
||||||
|
|
||||||
defer resp_payload.deinit();
|
|
||||||
|
|
||||||
while (true) {
|
|
||||||
const read = try reader.read(&buf);
|
|
||||||
try resp_payload.appendSlice(buf[0..read]);
|
|
||||||
if (read == 0) break;
|
|
||||||
}
|
|
||||||
const event_response = event_handler(req_allocator, resp_payload.items) catch |err| {
|
|
||||||
// Stack trace will return null if stripped
|
|
||||||
const return_trace = @errorReturnTrace();
|
|
||||||
if (return_trace) |rt|
|
|
||||||
log.err("Caught error: {}. Return Trace: {any}", .{ err, rt })
|
|
||||||
else
|
|
||||||
log.err("Caught error: {}. No return trace available", .{err});
|
|
||||||
const err_url = try std.fmt.allocPrint(req_allocator, "{s}{s}/runtime/invocation/{s}/error", .{ prefix, lambda_runtime_uri, req_id });
|
|
||||||
defer req_allocator.free(err_url);
|
|
||||||
const err_uri = try std.Uri.parse(err_url);
|
|
||||||
const content =
|
|
||||||
\\ {s}
|
|
||||||
\\ "errorMessage": "{s}",
|
|
||||||
\\ "errorType": "HandlerReturnedError",
|
|
||||||
\\ "stackTrace": [ "{any}" ]
|
|
||||||
\\ {s}
|
|
||||||
;
|
|
||||||
const content_fmt = if (return_trace) |rt|
|
|
||||||
try std.fmt.allocPrint(req_allocator, content, .{ "{", @errorName(err), rt, "}" })
|
|
||||||
else
|
|
||||||
try std.fmt.allocPrint(req_allocator, content, .{ "{", @errorName(err), "no return trace available", "}" });
|
|
||||||
defer req_allocator.free(content_fmt);
|
|
||||||
log.err("Posting to {s}: Data {s}", .{ err_url, content_fmt });
|
|
||||||
|
|
||||||
var err_headers = std.http.Headers.init(req_allocator);
|
|
||||||
defer err_headers.deinit();
|
|
||||||
err_headers.append(
|
|
||||||
"Lambda-Runtime-Function-Error-Type",
|
|
||||||
"HandlerReturned",
|
|
||||||
) catch |append_err| {
|
|
||||||
log.err("Error appending error header to post response for request id {s}: {}", .{ req_id, append_err });
|
|
||||||
std.os.exit(0);
|
|
||||||
continue;
|
|
||||||
};
|
|
||||||
var err_req = try client.request(.POST, err_uri, empty_headers, .{});
|
|
||||||
defer err_req.deinit();
|
|
||||||
err_req.start() catch |post_err| { // Well, at this point all we can do is shout at the void
|
|
||||||
log.err("Error posting response for request id {s}: {}", .{ req_id, post_err });
|
|
||||||
std.os.exit(0);
|
|
||||||
continue;
|
|
||||||
};
|
|
||||||
|
|
||||||
err_req.wait() catch |post_err| { // Well, at this point all we can do is shout at the void
|
|
||||||
log.err("Error posting response for request id {s}: {}", .{ req_id, post_err });
|
|
||||||
std.os.exit(0);
|
|
||||||
continue;
|
|
||||||
};
|
|
||||||
// TODO: Determine why this post is not returning
|
|
||||||
if (err_req.response.status != .ok) {
|
|
||||||
// Documentation says something about "exit immediately". The
|
|
||||||
// Lambda infrastrucutre restarts, so it's unclear if that's necessary.
|
|
||||||
// It seems as though a continue should be fine, and slightly faster
|
|
||||||
// std.os.exit(1);
|
|
||||||
log.err("Get fail: {} {s}", .{
|
|
||||||
@intFromEnum(err_req.response.status),
|
|
||||||
err_req.response.status.phrase() orelse "",
|
|
||||||
});
|
|
||||||
continue;
|
|
||||||
}
|
|
||||||
log.err("Post complete", .{});
|
|
||||||
continue;
|
|
||||||
};
|
|
||||||
// TODO: We should catch these potential alloc errors too
|
|
||||||
// TODO: This whole loop should be in another function so we can catch everything at once
|
|
||||||
const response_url = try std.fmt.allocPrint(req_allocator, "{s}{s}{s}/{s}/response", .{ prefix, lambda_runtime_uri, postfix, req_id });
|
|
||||||
defer req_allocator.free(response_url);
|
|
||||||
const response_uri = try std.Uri.parse(response_url);
|
|
||||||
const response_content = try std.fmt.allocPrint(req_allocator, "{s} \"content\": \"{s}\" {s}", .{ "{", event_response, "}" });
|
|
||||||
var resp_req = try client.request(.POST, response_uri, empty_headers, .{});
|
|
||||||
defer resp_req.deinit();
|
|
||||||
resp_req.transfer_encoding = .{ .content_length = response_content.len };
|
|
||||||
try resp_req.start();
|
|
||||||
try resp_req.writeAll(response_content); // TODO: AllocPrint + writeAll makes no sense
|
|
||||||
try resp_req.finish();
|
|
||||||
resp_req.wait() catch |err| {
|
|
||||||
// TODO: report error
|
|
||||||
log.err("Error posting response for request id {s}: {}", .{ req_id, err });
|
|
||||||
continue;
|
|
||||||
};
|
|
||||||
}
|
}
|
||||||
|
|
||||||
|
fn postResponse(self: Self, lambda_runtime_uri: []const u8, event_response: []const u8) !void {
|
||||||
|
const response_url = try std.fmt.allocPrint(
|
||||||
|
self.allocator,
|
||||||
|
"{s}{s}{s}/{s}/response",
|
||||||
|
.{ prefix, lambda_runtime_uri, postfix, self.request_id },
|
||||||
|
);
|
||||||
|
defer self.allocator.free(response_url);
|
||||||
|
const response_uri = try std.Uri.parse(response_url);
|
||||||
|
var cl = std.http.Client{ .allocator = self.allocator };
|
||||||
|
defer cl.deinit();
|
||||||
|
var req = try cl.request(.POST, response_uri, empty_headers, .{});
|
||||||
|
// var req = try client.?.request(.POST, response_uri, empty_headers, .{});
|
||||||
|
defer req.deinit();
|
||||||
|
const response_content = try std.fmt.allocPrint(
|
||||||
|
self.allocator,
|
||||||
|
"{{ \"content\": \"{s}\" }}",
|
||||||
|
.{event_response},
|
||||||
|
);
|
||||||
|
defer self.allocator.free(response_content);
|
||||||
|
|
||||||
|
req.transfer_encoding = .{ .content_length = response_content.len };
|
||||||
|
try req.start();
|
||||||
|
try req.writeAll(response_content);
|
||||||
|
try req.finish();
|
||||||
|
try req.wait();
|
||||||
|
}
|
||||||
|
};
|
||||||
|
|
||||||
|
fn getEvent(allocator: std.mem.Allocator, event_data_uri: std.Uri) !?Event {
|
||||||
|
// TODO: There is something up with using a shared client in this way
|
||||||
|
// so we're taking a perf hit in favor of stability. In a practical
|
||||||
|
// sense, without making HTTPS connections (lambda environment is
|
||||||
|
// non-ssl), this shouldn't be a big issue
|
||||||
|
var cl = std.http.Client{ .allocator = allocator };
|
||||||
|
defer cl.deinit();
|
||||||
|
var req = try cl.request(.GET, event_data_uri, empty_headers, .{});
|
||||||
|
// var req = try client.?.request(.GET, event_data_uri, empty_headers, .{});
|
||||||
|
// defer req.deinit();
|
||||||
|
|
||||||
|
try req.start();
|
||||||
|
try req.finish();
|
||||||
|
// Lambda freezes the process at this line of code. During warm start,
|
||||||
|
// the process will unfreeze and data will be sent in response to client.get
|
||||||
|
try req.wait();
|
||||||
|
if (req.response.status != .ok) {
|
||||||
|
// Documentation says something about "exit immediately". The
|
||||||
|
// Lambda infrastrucutre restarts, so it's unclear if that's necessary.
|
||||||
|
// It seems as though a continue should be fine, and slightly faster
|
||||||
|
// std.os.exit(1);
|
||||||
|
log.err("Lambda server event response returned bad error code: {} {s}", .{
|
||||||
|
@intFromEnum(req.response.status),
|
||||||
|
req.response.status.phrase() orelse "",
|
||||||
|
});
|
||||||
|
return error.EventResponseNotOkResponse;
|
||||||
|
}
|
||||||
|
|
||||||
|
var request_id: ?[]const u8 = null;
|
||||||
|
var content_length: ?usize = null;
|
||||||
|
for (req.response.headers.list.items) |h| {
|
||||||
|
if (std.ascii.eqlIgnoreCase(h.name, "Lambda-Runtime-Aws-Request-Id"))
|
||||||
|
request_id = h.value;
|
||||||
|
if (std.ascii.eqlIgnoreCase(h.name, "Content-Length")) {
|
||||||
|
content_length = std.fmt.parseUnsigned(usize, h.value, 10) catch null;
|
||||||
|
if (content_length == null)
|
||||||
|
log.warn("Error parsing content length value: '{s}'", .{h.value});
|
||||||
|
}
|
||||||
|
// TODO: XRay uses an environment variable to do its magic. It's our
|
||||||
|
// responsibility to set this, but no zig-native setenv(3)/putenv(3)
|
||||||
|
// exists. I would kind of rather not link in libc for this,
|
||||||
|
// so we'll hold for now and think on this
|
||||||
|
// if (std.mem.indexOf(u8, h.name.value, "Lambda-Runtime-Trace-Id")) |_|
|
||||||
|
// std.process.
|
||||||
|
// std.os.setenv("AWS_LAMBDA_RUNTIME_API");
|
||||||
|
}
|
||||||
|
if (request_id == null) {
|
||||||
|
// We can't report back an issue because the runtime error reporting endpoint
|
||||||
|
// uses request id in its path. So the best we can do is log the error and move
|
||||||
|
// on here.
|
||||||
|
log.err("Could not find request id: skipping request", .{});
|
||||||
|
return null;
|
||||||
|
}
|
||||||
|
if (content_length == null) {
|
||||||
|
// We can't report back an issue because the runtime error reporting endpoint
|
||||||
|
// uses request id in its path. So the best we can do is log the error and move
|
||||||
|
// on here.
|
||||||
|
log.err("No content length provided for event data", .{});
|
||||||
|
return null;
|
||||||
|
}
|
||||||
|
const req_id = request_id.?;
|
||||||
|
log.debug("got lambda request with id {s}", .{req_id});
|
||||||
|
|
||||||
|
var resp_payload = try std.ArrayList(u8).initCapacity(allocator, content_length.?);
|
||||||
|
defer resp_payload.deinit();
|
||||||
|
try resp_payload.resize(content_length.?);
|
||||||
|
var response_data = try resp_payload.toOwnedSlice();
|
||||||
|
errdefer allocator.free(response_data);
|
||||||
|
_ = try req.readAll(response_data);
|
||||||
|
|
||||||
|
return Event.init(
|
||||||
|
allocator,
|
||||||
|
response_data,
|
||||||
|
try allocator.dupe(u8, req_id),
|
||||||
|
);
|
||||||
}
|
}
|
||||||
|
|
||||||
////////////////////////////////////////////////////////////////////////
|
////////////////////////////////////////////////////////////////////////
|
||||||
|
@ -320,7 +403,7 @@ fn test_run(allocator: std.mem.Allocator, event_handler: HandlerFn) !std.Thread
|
||||||
);
|
);
|
||||||
}
|
}
|
||||||
|
|
||||||
fn lambda_request(allocator: std.mem.Allocator, request: []const u8) ![]u8 {
|
fn lambda_request(allocator: std.mem.Allocator, request: []const u8, request_count: usize) ![]u8 {
|
||||||
var arena = std.heap.ArenaAllocator.init(allocator);
|
var arena = std.heap.ArenaAllocator.init(allocator);
|
||||||
defer arena.deinit();
|
defer arena.deinit();
|
||||||
var aa = arena.allocator();
|
var aa = arena.allocator();
|
||||||
|
@ -344,7 +427,7 @@ fn lambda_request(allocator: std.mem.Allocator, request: []const u8) ![]u8 {
|
||||||
// booleans to know when to shut down. This function is designed for a
|
// booleans to know when to shut down. This function is designed for a
|
||||||
// single request/response pair only
|
// single request/response pair only
|
||||||
|
|
||||||
lambda_remaining_requests = 1; // in case anyone messed with this, we will make sure we start
|
lambda_remaining_requests = request_count;
|
||||||
server_remaining_requests = lambda_remaining_requests.? * 2; // Lambda functions
|
server_remaining_requests = lambda_remaining_requests.? * 2; // Lambda functions
|
||||||
// fetch from the server,
|
// fetch from the server,
|
||||||
// then post back. Always
|
// then post back. Always
|
||||||
|
@ -379,7 +462,25 @@ test "basic request" {
|
||||||
const expected_response =
|
const expected_response =
|
||||||
\\{ "content": "{"foo": "bar", "baz": "qux"}" }
|
\\{ "content": "{"foo": "bar", "baz": "qux"}" }
|
||||||
;
|
;
|
||||||
const lambda_response = try lambda_request(allocator, request);
|
const lambda_response = try lambda_request(allocator, request, 1);
|
||||||
|
defer deinit();
|
||||||
|
defer allocator.free(lambda_response);
|
||||||
|
try std.testing.expectEqualStrings(expected_response, lambda_response);
|
||||||
|
}
|
||||||
|
|
||||||
|
test "several requests do not fail" {
|
||||||
|
// std.testing.log_level = .debug;
|
||||||
|
const allocator = std.testing.allocator;
|
||||||
|
const request =
|
||||||
|
\\{"foo": "bar", "baz": "qux"}
|
||||||
|
;
|
||||||
|
|
||||||
|
// This is what's actually coming back. Is this right?
|
||||||
|
const expected_response =
|
||||||
|
\\{ "content": "{"foo": "bar", "baz": "qux"}" }
|
||||||
|
;
|
||||||
|
const lambda_response = try lambda_request(allocator, request, 5);
|
||||||
|
defer deinit();
|
||||||
defer allocator.free(lambda_response);
|
defer allocator.free(lambda_response);
|
||||||
try std.testing.expectEqualStrings(expected_response, lambda_response);
|
try std.testing.expectEqualStrings(expected_response, lambda_response);
|
||||||
}
|
}
|
||||||
|
|
Loading…
Reference in New Issue
Block a user