Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Convert null to NA and vice versa #32

Open
wants to merge 1 commit into
base: develop
Choose a base branch
from
Open

Convert null to NA and vice versa #32

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
328 changes: 234 additions & 94 deletions main.js
View file
Open in desktop

Large diffs are not rendered by default.

328 changes: 234 additions & 94 deletions rserve.js
View file
Open in desktop

Large diffs are not rendered by default.

3 changes: 2 additions & 1 deletion rserve.min.js
View file
Open in desktop

Large diffs are not rendered by default.

15 changes: 11 additions & 4 deletions src/parse.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,12 +60,16 @@ function read(m)
read_int_vector: function(length) {
var old_offset = this.offset;
this.offset += length;
return this.msg.make(Int32Array, old_offset, length);
return _.map(this.msg.make(Int32Array, old_offset, length), function(v) {
return v !== Rserve.Rsrv.INTEGER_NA ? v : null;
});
},
read_double_vector: function(length) {
var old_offset = this.offset;
this.offset += length;
return this.msg.make(Float64Array, old_offset, length);
return _.map(this.msg.make(Float64Array, old_offset, length), function(v) {
return !Rserve.Rsrv.IS_DOUBLE_NA(v) ? v : null;
});
},

//////////////////////////////////////////////////////////////////////
Expand All @@ -84,7 +88,10 @@ function read(m)
var current_str = "";
for (var i=0; i<a.length; ++i)
if (a[i] === 0) {
current_str = decodeURIComponent(escape(current_str));
if (current_str.length !== 1 || current_str.charCodeAt(0) !== Rserve.Rsrv.STRING_NA)
current_str = decodeURIComponent(escape(current_str));
else
current_str = null;
result.push(current_str);
current_str = "";
} else {
Expand All @@ -96,7 +103,7 @@ function read(m)
var l2 = this.read_int();
var s = this.read_stream(length-4);
var a = _.map(s.make(Uint8Array).subarray(0, l2), function(v) {
return v ? true : false;
return v !== Rserve.Rsrv.BOOL_NA ? v ? true : false : null;
});
return [Rserve.Robj.bool_array(a, attributes), length];
},
Expand Down
26 changes: 13 additions & 13 deletions src/robj.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,20 +55,20 @@ Rserve.Robj = {
if (_.isUndefined(this.attributes)) {
return values;
} else {
// FIXME: there is no reason why names should be the first or only
// attribute, so the code should really look
// for "names" and not cry if it doesn't exist
// FIXME: there is no reason why names should be the first or only
// attribute, so the code should really look
// for "names" and not cry if it doesn't exist
if (this.attributes.value[0].name == "names") {
var keys = this.attributes.value[0].value.value;
var result = {};
_.each(keys, function(key, i) {
result[key] = values[i];
result[key] = values[i];
});
return result;
}
// FIXME: how can we pass other important attributes
// like "class" ?
return values;
}
// FIXME: how can we pass other important attributes
// like "class" ?
return values;
}
}
}),
Expand All @@ -84,11 +84,11 @@ Rserve.Robj = {
if (_.isUndefined(this.attributes)) {
return values;
} else {
// FIXME: lang doens't have "names" attribute since
// names are sent as tags (langs are pairlists)
// so this seems superfluous (it is dangerous
// if lang ever had attributes since there is
// no reason to fail in that case)
// FIXME: lang doens't have "names" attribute since
// names are sent as tags (langs are pairlists)
// so this seems superfluous (it is dangerous
// if lang ever had attributes since there is
// no reason to fail in that case)
if(this.attributes.value[0].name!="names")
throw "expected names here";
var keys = this.attributes.value[0].value.value;
Expand Down
4 changes: 2 additions & 2 deletions src/rserve.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -324,8 +324,8 @@ Rserve.create = function(opts) {
closeFile: function(k) {
_cmd(Rserve.Rsrv.CMD_closeFile, new ArrayBuffer(0), k, "");
},
set: function(key, value, k) {
_cmd(Rserve.Rsrv.CMD_setSEXP, [_encode_string(key), _encode_value(value)], k, "");
set: function(key, value, k, forced_type) {
_cmd(Rserve.Rsrv.CMD_setSEXP, [_encode_string(key), _encode_value(value, forced_type)], k, "");
},

//////////////////////////////////////////////////////////////////////
Expand Down
33 changes: 33 additions & 0 deletions src/rsrv.js
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,26 @@
// Simple constants and functions are defined here,
// in correspondence with Rserve's Rsrv.h

function custom_nan(int32_words) {
// DataView.getFloat64() canonicalizes any NaNs with custom bit patterns, so
// instead use what is essentially a C union in JavaScript. Unfortunately,
// that means we have to deal with endianness.
var buf = new ArrayBuffer(8);
var intRep = new Int32Array(buf);
var floatRep = new Float64Array(buf);

// Server may be little endian.
intRep[0] = int32_words[1];
intRep[1] = int32_words[0];
if (isNaN(floatRep[0]))
return floatRep[0];

// Server may be big endian.
intRep[0] = int32_words[0];
intRep[1] = int32_words[1];
return floatRep[0];
}

Rserve.Rsrv = {
PAR_TYPE: function(x) { return x & 255; },
PAR_LEN: function(x) { return x >>> 8; },
Expand Down Expand Up @@ -112,10 +132,23 @@ Rserve.Rsrv = {
BOOL_FALSE : 0,
BOOL_NA : 2,

// See Arith.h in R.
INTEGER_NA : (2147483647 + 1) | 0,
STRING_NA : 0xff,
DOUBLE_NA : custom_nan([ 0x7ff00000, 0x000007a2 ]),

GET_XT: function(x) { return x & 63; },
GET_DT: function(x) { return x & 63; },
HAS_ATTR: function(x) { return (x & Rserve.Rsrv.XT_HAS_ATTR) > 0; },
IS_LARGE: function(x) { return (x & Rserve.Rsrv.XT_LARGE) > 0; },
IS_DOUBLE_NA: function(x) {
var view = new DataView(new ArrayBuffer(16));
view.setFloat64(0, x);
view.setFloat64(8, Rserve.Rsrv.DOUBLE_NA);
// Any operations (not involving NaN) still produce NA, but they convert
// the signaling NaN to quiet NaN by toggling the highest mantissa bit.
return (view.getInt32(0) & ~0x00080000) === view.getInt32(8) && view.getInt32(4) === view.getInt32(12);
},

// # FIXME A WHOLE LOT OF MACROS HERE WHICH ARE PROBABLY IMPORTANT
// ##############################################################################
Expand Down
65 changes: 45 additions & 20 deletions src/write.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,13 +29,15 @@ Rserve.type_id = function(value)
if (!_.isUndefined(value.byteLength) && !_.isUndefined(value.slice))
return Rserve.Rsrv.XT_RAW;

// before coercion, NA is R is a logical, so if entire array is null, then
// translate it into a logical vector accordingly
if (_.isArray(value) && _.all(value, function(el) { return _.isNull(el) || _.isBoolean(el); }))
return Rserve.Rsrv.XT_ARRAY_BOOL;

// lists of strings (important for tags)
if (_.isArray(value) && _.all(value, function(el) { return typeof el === 'string'; }))
if (_.isArray(value) && _.all(value, function(el) { return _.isNull(el) || _.isString(el); }))
return Rserve.Rsrv.XT_ARRAY_STR;

if (_.isArray(value) && _.all(value, function(el) { return typeof el === 'boolean'; }))
return Rserve.Rsrv.XT_ARRAY_BOOL;

// arbitrary lists
if (_.isArray(value))
return Rserve.Rsrv.XT_VECTOR;
Expand Down Expand Up @@ -71,23 +73,31 @@ Rserve.determine_size = function(value, forced_type)
case Rserve.Rsrv.XT_NULL:
return final_size(0);
case Rserve.Rsrv.XT_ARRAY_BOOL:
if (_.isBoolean(value))
if (_.isNull(value) || _.isBoolean(value))
return final_size(8);
else
return final_size((value.length + 7) & ~3);
case Rserve.Rsrv.XT_ARRAY_STR:
if (_.isArray(value))
return final_size(_.reduce(value, function(memo, str) {
// FIXME: this is a bit silly, since we'll be re-encoding this twice: once for the size and second time for the content
var utf8 = unescape(encodeURIComponent(str));
// FIXME: this is a bit silly, since we'll be re-encoding this twice: once for the size and second time for the content
var utf8;
if (_.isNull(str))
utf8 = String.fromCharCode(Rserve.Rsrv.STRING_NA);
else
utf8 = unescape(encodeURIComponent(str));
return memo + utf8.length + 1;
}, 0));
else {
var utf8 = unescape(encodeURIComponent(value));
var utf8;
if (_.isNull(value))
utf8 = String.fromCharCode(Rserve.Rsrv.STRING_NA);
else
utf8 = unescape(encodeURIComponent(value));
return final_size(utf8.length + 1);
}
}
case Rserve.Rsrv.XT_ARRAY_DOUBLE:
if (_.isNumber(value))
if (_.isNull(value) || _.isNumber(value))
return final_size(8);
else
return final_size(8 * value.length);
Expand Down Expand Up @@ -143,37 +153,52 @@ Rserve.write_into_view = function(value, array_buffer_view, forced_type, convert
case Rserve.Rsrv.XT_NULL:
break;
case Rserve.Rsrv.XT_ARRAY_BOOL:
if (_.isBoolean(value)) {
if (_.isNull(value)) {
write_view.setInt32(payload_start, 1);
write_view.setInt8(payload_start + 4, Rserve.Rsrv.BOOL_NA);
} else if (_.isBoolean(value)) {
write_view.setInt32(payload_start, 1);
write_view.setInt8(payload_start + 4, value ? 1 : 0);
write_view.setInt8(payload_start + 4, value ? Rserve.Rsrv.BOOL_TRUE : Rserve.Rsrv.BOOL_FALSE);
} else {
write_view.setInt32(payload_start, value.length);
for (i=0; i<value.length; ++i)
write_view.setInt8(payload_start + 4 + i, value[i] ? 1 : 0);
write_view.setInt8(payload_start + 4 + i, !_.isNull(value[i]) ? value[i] ? Rserve.Rsrv.BOOL_TRUE : Rserve.Rsrv.BOOL_FALSE : Rserve.Rsrv.BOOL_NA);
}
break;
case Rserve.Rsrv.XT_ARRAY_STR:
if (_.isArray(value)) {
if (_.isNull(value)) {
write_view.setUint8(payload_start, Rserv.Rsrv.STRING_NA);
write_view.setUint8(payload_start + 1, 0);
} else if (_.isArray(value)) {
var offset = payload_start;
_.each(value, function(el) {
var utf8 = unescape(encodeURIComponent(el));
for (var i=0; i<utf8.length; ++i, ++offset)
write_view.setUint8(offset, utf8.charCodeAt(i));
if (_.isNull(el)) {
write_view.setUint8(offset++, Rserv.Rsrv.STRING_NA);
} else {
var utf8 = unescape(encodeURIComponent(el));
for (var i=0; i<utf8.length; ++i, ++offset)
write_view.setUint8(offset, utf8.charCodeAt(i));
}
write_view.setUint8(offset++, 0);
});
} else {
var utf8 = unescape(encodeURIComponent(value));
var utf8 = unescape(encodeURIComponent(value));
for (i=0; i<utf8.length; ++i)
write_view.setUint8(payload_start + i, utf8.charCodeAt(i));
write_view.setUint8(payload_start + utf8.length, 0);
}
break;
case Rserve.Rsrv.XT_ARRAY_DOUBLE:
if (_.isNumber(value))
if (_.isNull(value)) {
write_view.setFloat64(payload_start, Rserv.Rsrv.DOUBLE_NA);
} else if (_.isNumber(value))
write_view.setFloat64(payload_start, value);
else {
for (i=0; i<value.length; ++i)
write_view.setFloat64(payload_start + 8 * i, value[i]);
if (_.isNull(value[i]))
write_view.setFloat64(payload_start + 8 * i, Rserv.Rsrv.DOUBLE_NA);
else
write_view.setFloat64(payload_start + 8 * i, value[i]);
}
break;
case Rserve.Rsrv.XT_RAW:
Expand Down