The engine keeps Java's packet capture (capture.ts): per network card
every frame it sent (when it leaves on the cable) or received (also one
for another card), once, with time and number; switches log nothing;
frames the packet loss dropped are kept apart; the capture outlives a
stop and a new start, as in Java, until clearCapture. Per port instead
of per MAC, at most 10,000 frames a card.
expectSameCaptures compares each card's capture with Java's list of that
card in the ARP, WLAN, DNS, TCP, NAT, RIP and single-client DHCP runs.
It found that ScenarioNetwork ignored a switch's learning time (Java ran
300 s, the core the scenarios' 5 s); the oracle sets it now, and the TCP
and DHCP runs are recorded again. A core test found that a new start
numbers frames from 1 again and the capture took them as seen.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
WlanRun pings on the new network wlan-lan: over wireless cables, over a
plain cable, and to and from a notebook on the switch's SSID without a
cable. The engine matches Java without a change: it runs the links only.
Swift's wirelessAssociations, which join a host to its SSID's switch at
run time, are not taken. ScenarioNetwork sets wireless cables, wireless
cards and SSIDs; a corpus test pings over the wireless cables of
multicast_server.fls.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The core translates before the IP thread, as Java's NAT threads sit
between card and Ethernet: out, the WAN address and the entry's port (ICMP
keeps its identifier); in, by remote end, port and protocol, else by the
port forwarding, which then makes a dynamic entry; a ping at the WAN
address is answered by the router, packets from outside for the LAN are
dropped, both sides pass the router's firewall, TTL 1 going out gets time
exceeded. Dynamic entries expire at Java's check each second, 300 s times
the delay factor old (55 minutes at 90 %), refreshed only by what comes in.
The home router forwards as a router does.
Measured and not taken: when a LAN endpoint sends to a second remote end,
Java gives it the entry's WAN port but no entry, and the answer stays at
the router; the core adds an entry for each new remote end on that port.
The oracle now records every card of a node (the second card of routers
was missing); the ARP and RIP runs are recorded anew and still match.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ScenarioNetwork builds home routers (their first card the WAN port) and
their port forwarding (portForwarding, through GatewayFirmware.setStaticNAT).
NatRun plays six stories on nat-home's network and writes the frames, the
calls and the router's NAT table: a ping, a UDP exchange, the same socket
to a second server (whose answer never reaches the LAN: Java keeps one
entry per LAN endpoint, for the first remote only), a TCP connection out,
a forwarded one in, and pings at the WAN address and past it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RipRun runs rip-chain's three routers in Java FILIUS, each router's random
set by reflection to the core's ripSeed, and writes the frames, every
change of each RIP table and the calls. converge: the tables fill, a ping
at 45 s. expire: R3 falls silent at 10 s; R2 marks its network 16 hops only
at its own next beacon, R1 passes it on at once, the ping gets ICMP 3/0.
Both match: per sending card the same messages at the same times, every
table through the same states. The first 5 s are compared by the state the
tables reach, since Java's routers start one after the other.
Measured on the JDK: Java's pause divides nextFloat() by 3 in float; the
core does too now (a millisecond off in one draw of about 10,000 before).
Scenarios name their RIP routers in the new field rip.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Where the Swift template departs from Java's filius.software.rip, Java's
code rules: a learned route gets 16 hops at the router's next beacon after
it expired (RIPTable.check), not when it expires; messages count from any
source port; each router draws its pauses from a random of its own (Java:
an unseeded Random per router), which the core seeds from the scenario and
the router's id (ripSeed), so that the Java oracle can give it the same.
A route of hop count 0 to each own network, the first beacon after a
second, then every 25.2 to 35.2 s; a message per interface without the
routes through it; a new route or a longer one from the same router sends
at once; forwarding by the table alone, fewest hops before longest mask.
Messages are read with Java's parser (javaIpCheck, javaParseInt, moved from
fls to core for it).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The resolver runs on the events now (a query, its dnsQueryTimeout, its
answer at the socket); the blocking resolveDns waits for it, and a
recursive server uses it: its own records, else the zone's name server
from its records, else its own DNS server, else a referral
(DNSServer.answerWithRemoteData).
DnsRun plays five stories on the new dns-recursive network. Measured and
not taken: a server that is its zone's name server asks itself, which asks
itself again, without end (the client times out, then every level floods
its own DNS server); Java guards only the step through its own DNS server.
The core asks itself in neither step. That story is compared up to the
client's query (JAVA_UNTIL).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Every packet that comes in goes through its node's IP thread (UDP, TCP) or
ICMP thread (ICMP), as Java's IPThread and ICMPThread: deliver, answer an
echo request or forward, one after the other; while one waits for ARP, the
next wait behind it. A packet a node sends to itself is delivered at once,
as Java's IP.senden does. The transport send threads and these are one
mechanism now (handToProtocolThread).
ArpRun has three stories through the router (routedUdpBehind,
routedPingBesideUdp, routedPingBehindPing); the oracle writes ICMP frames
with type, code, addresses and sequence number. All seven ArpRun stories
match Java FILIUS.
Measured and not taken: Java's ARP wakes one waiter with notify, not the
one whose answer came; the core wakes all that wait for the address.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
TcpRun loses exactly one frame (loseFirst: the card's output buffer drops
it; the sender's packet log still shows it, as with setDrop) in three new
stories: lostSynAck, lostFin, lostLastFin.
Measured: Java never repeats a SYN+ACK or a FIN. The server gives a
handshake up after one timeout (3 RTT) and closes with a FIN from SYN_RCVD
(its boundary check sometimes lets it wait twice; it means once). After its
FIN a socket waits at most five times, each wait ended by a segment or the
timeout; then it is gone. The core does the same (TCP_CLOSING_WAITS, with
TIME_WAIT in the loop) and shows CLOSED where Java leaves the state
standing (a named Java bug).
One departure: a repeated SYN in SYN_RCVD gets the SYN+ACK once more and
the wait starts anew; Java closes there, but its client never repeats the
SYN. lostSynAck is compared up to the second SYN (JAVA_UNTIL); the rest is
an engine test.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A send hands its frame to the card and returns, as in Java; the frames
travel while the clock advances. Swift's send ran the network until every
frame had arrived; the option swiftSends keeps that for the Swift
comparison. Waiting calls (TCP connect, acknowledgement) now run event by
event until the answer instead of to the timeout.
UDP and TCP send through a thread per node and protocol (Java's
TransportProtokoll): one segment after the other; while IP waits for ARP,
the next send blocks and a segment from an event queues behind it. Senders
that wait take their turns in the order they came (a fair lock; Java's
leaves the order to chance).
All four ArpRun stories now match Java FILIUS, including when each call
returns. Tests let frames arrive explicitly with letFramesArrive.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
An ARP lookup asks, waits a round trip, asks once more and gives up after
another round trip; then host unreachable, observed at the sender itself or
sent back by the router (for ICMP only for an echo request). A waiting
sender wakes after what else happens at that time. A ping then waits a
round trip for its answer, counted from when it left. Swift asks twice at
once while events run and lets the packet wait without end when the
address exists anywhere in the network; the option swiftArp keeps that for
the Swift comparison, and three scenarios are compared with Swift until the
trace where ARP fails or a ping goes unanswered.
arpJava.test.ts compares ArpRun: both pings match; the two UDP stories wait
for Java's transport thread (next step). DNS without a server now matches
Java too.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Sends to addresses no card has on the new arp-routed network (ping on the
LAN and behind a router, two UDP datagrams, one behind a failing one):
frames and calls. ScenarioNetwork builds routers and their interfaces.
Measured: ARP asks twice, a round trip each, then host unreachable; the
thread that waits is the ping's caller, the router's IP thread, or the
node's transport thread per protocol, which blocks the next send.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
DNS follows Java's DNSServer and Resolver, not Swift's typed format:
Java's text messages (dns.ts), a server on port 53 answering from its
records (no question in the answer, RCODE 0, a referral with NS and A in
the answer section), an iterative resolver that waits five round trips per
query and remembers only the servers it was referred to. Recursion throws
NotPortedError until the next step. The snapshot carries the DNS servers
switched on. receiveUdp returns as soon as a datagram is there, as Java's
empfangen does.
dnsJava.test.ts compares DnsRun's five lookups; four match frame by frame,
the missing server waits for Java's ARP (named as not yet like Java).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A resolver's lookup on the new dns-lan network (A, MX, a referral by NS,
an unknown name, a missing server): every frame and the call's result.
ScenarioNetwork installs a scenario's DNS servers. The Swift comparison
reads only scenarios with a Swift result.
Measured: Java's text messages (ID=... QR=... header, question and record
lines), no question in the answer, RCODE 0 always, a referral in the answer
section, an iterative resolver waiting 5 RTT per query.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Java's SimplexVerbindung sleeps a cable's delay with the port's buffer
locked: frames on one cable leave one after the other, and a frame shows
in the packet log when it leaves. The engine now queues a frame in the
output buffer until the one before has arrived (ethernetLinkDeparture).
Swift sends a cable's frames at once; the option swiftCables keeps that for
the Swift comparison, which goes on checking everything else trace by
trace.
With the queue, the Java comparisons match within 120 ms (was 200); a
state of the TCP stories within 250 ms, where the engine's calls block
one after the other.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
TCP follows Java's TCPSocket where it differs from Swift: each attempt and
TIME_WAIT wait three round trips that follow the speed (8250 ms at 90 %),
counted from when the segment is handed down, before ARP; the next segment
ends TIME_WAIT; the other end closes by itself on FIN; data no one
acknowledges closes with FIN; the accepted socket has the listener's
initial sequence number, and the listener takes the next.
Not taken from Java: three SYNs, not one (Java's retry loop ends after
the first timeout), named in tcpJava.test.ts as JAVA_BUGS.
tcpJava.test.ts compares TcpRun's three stories frame by frame and state
by state; the Swift comparison only checks that the TCP scenarios differ.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Three stories of socket calls on tcp-lan's network (exchange, closedPort,
lostData): every frame with TCP flags and numbers, when each call started,
returned or failed, and each socket's states. Network building and the
packet log move from DhcpRun to ScenarioNetwork.
Measured: 3 RTT per attempt (8250 ms at 90 %), an orderly close after a
failed send, TIME_WAIT of 3 RTT, the passive side closing by itself on FIN,
one SYN only (Java's retry loop ends after the first timeout).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The client follows Java's DHCPClient instead of Swift: one DISCOVER or
REQUEST per attempt, an answer that does not fit is read and dropped, the
ARP check waits a round trip (an answer for the address ends it at once,
then DECLINE), a timeout repeats the state's message, and the fifth turns
to a link-local address 169.254.1.x/16 (ARP probe, ARP reply to all); no
old address is given back. DHCP round trip times follow the speed as in
Java (50 cables, 2750 ms at 90 %; Swift: fixed 2500 ms).
Not taken from Java: the client gives up after five occupied link-local
addresses (Java searches on without end), and an ARP packet for another
address does not end the check.
dhcpJava.test.ts compares the four DHCP scenarios with DhcpRun's Java
runs; the Swift comparison only checks that they differ.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Builds a core scenario's network in Java FILIUS, starts it as the action
mode does and writes every frame each card saw (the Lauscher) and the
addresses at the end. References for the four DHCP scenarios: one
DISCOVER/REQUEST per attempt, an ARP check that waits one RTT, DECLINE on a
conflict, link-local 169.254.1.x after the fifth failure.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
engine.ts ports Swift's "DHCP server and automatic client": an active
server listens on port 67 from the start and offers a static assignment
before its range, round from the address after the last one offered;
offers last 10 s, dynamic leases a day, and a server whose offer another
server beat blacklists that address for a while. A client starts as an
event: DISCOVER, OFFER, a check by ARP, DECLINE, REQUEST, ACK or NAK,
2.5 s per wait, and after ten failures its old address back. The snapshot
carries the DHCP settings of clients and servers.
Four scenarios match the Swift engine. One Swift error is named in the
oracle test: a client run directly is advanced a second time, to FINISH,
by the server's answer within its own DISCOVER, and the outer run then
writes back its stale copy of the status (Swift's value types); the
address and every trace agree, only the reported status is wrong in
Swift. Swift's three DHCP engine tests are ported, with tests for
stopping and the message format. 15 mutation probes, 14 caught; the other
concerns the home router, which runs only once NAT is ported.
Two departures from Java remain, to be fixed next: from the event queue
the client's ARP check does not wait, so an address conflict is never
found, and each answer that wakes the client makes it send its state's
message again (a second DISCOVER and REQUEST, the second answered with
NAK). Java waits a round trip and sends once per attempt.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
engine.ts ports Swift's "Java-compatible TCP sockets" in one piece, since
half a state machine could not be checked: server and client sockets (a
free port from Java's random numbers on their own seed), the three-way
handshake with its result (connected, invalid socket, unreachable, timed
out), segments of 1460 bytes, messages up to the push, a timeout of
3 × 2.5 s with at most three attempts, closing through every Java state
including TIME_WAIT (2.5 s), sequence numbers modulo 2^32.
Three scenarios match the Swift engine trace for trace: a LAN with data
over three segments, closing from both ends, a closed port and an
unreachable host; two networks through a router with a firewall rule,
data lost until the connection closes, and an accepted socket that does
not hold the port of its closed listener; CLOSING, reached by a lost FIN.
The engine delivers a FIN within the same close, so both ends close at
once only with packet loss; an empty message does not arrive in Swift (it
does not advance the sequence number), nor here.
Swift's four TCP engine tests are ported, with tests for stopping, a
foreign SYN+ACK at a listener and the flag names. 17 mutation probes, 16
caught; the other (delivered, not waited, not connected gives unreachable)
cannot be reached through Swift's engine, which never sends a reset and
always acknowledges correctly.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
engine.ts ports the UDP part of Swift's "UDP sockets and Router RIP": a
socket bound to a given port or to a free one drawn from 49152–65534 with
Java's random numbers on Swift's seed; sending, also as a limited broadcast
on every interface (or the bound one) with TTL 1; delivery into the queue
of every socket on the destination port, which then answers to the last
sender; receiving, with a timeout that advances the virtual clock; closing.
Stopping forgets sockets and queues. RIP stays not ported; Swift's DHCP
handling of a datagram has nothing to do until DHCP is in the snapshot.
The oracle gets the steps bindUdp, sendUdp, receiveUdp and closeSocket on
both sides and reports the sockets. Two new scenarios — a LAN, and two
networks through a router with a UDP firewall rule and a broadcast that
does not cross it — match the Swift engine trace for trace, the random
ports included. Swift's four UDP engine tests are ported, with a test for
stopping. 11 mutation probes were all caught (one only after the stop test
was added).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The model carries the captured beans of a project's own applications
(ownApplicationsByNodeId); the export writes them again on the system
software of their node, and their classes go with the archive's other
entries. Checked by Java: from the exported files FILIUS loads
Accumulator and Broadcast as from the original; the export test no longer
leaves own applications out, and a probe that drops them is caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A project's own applications (classes under projekt/anwendungen, installed
on a host) are no longer left out. The decoder now keeps the Java type of
every primitive — at setters, constructor and call arguments, also through
shared ids like Integer0 — and which properties an object's own getter
made. beanCapture.ts turns a decoded bean back into what the writer writes,
with the node's system software as a mark the writer replaces by the
system software it writes; what the format cannot say is refused, not
guessed (an unresolved value, a Field, a factory-made object, an object a
getter reaches elsewhere in the file). The reader keeps own applications as
captured beans; the corpus no longer warns.
The capture first took a setter with another object's getter result for an
own getter; the round-trip test found it. The Java oracle did not show
whether an application runs on its node's system software, so a mutation
writing the mark instead survived; it now reports onItsSystem (true for all
192 applications of the corpus).
Checked by Java: from the files written again, FILIUS loads Accumulator and
Broadcast installed, switched on and on their node's system software, as
in the original. The round-trip test covers every primitive, null, a
class, an array, shared objects, getter chains, constructed elements, map
reads and static constants.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
FlsOracle now loads through SzenarioVerwaltung.laden, FILIUS's full
loading path: the archive unpacked, the project's own applications copied
where FILIUS's class loader finds them, then the configuration.
--network keeps the configuration-only load (netzwerkLaden).
On the corpus this changes two files: anwendung_komplett.fls and
multicast_server.fls then have their own applications (Accumulator,
Broadcast) installed. Java's views of the written and the exported files
are byte-equal to before: Java-FILIUS 2.14 opens our exports, with the
classes they carry, as it opens a file of its own, and sees what it sees
in the original. Until own applications on hosts are written (B2), the
reader, writer and export tests leave them out (test/ownApplications.ts);
a test checks that only those two files are touched.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The archive's entries besides the configuration — the project's own Java
applications under projekt/anwendungen (.java, .class,
EigeneAnwendungen.txt) — now go through the editor model:
openProjectArchive reads a whole .fls into the model and keeps them in its
archiveEntries, saveProjectArchive writes a whole .fls from the model with
them unchanged. They cannot run here, but must not be lost.
Checked on every corpus file: each entry comes back byte for byte (7 files
carry applications of their own); a probe that drops them is caught.
Installing such applications on a host is the next step; Java loads those
only through its full loading path, which the oracle is to use first.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
exportProject.ts, step X4b, which completes the export: the e-mail beans
from the model — the server's domain, accounts and mailboxes, the client's
three folders and address book. The account files are written anew from
the model's accounts, so that they say what the model says:
/mailserver/konten.txt, which Java's server reads at every simulation
start, and /konten.txt, the client's account. The format is FILIUS 2.14's
(listeBenutzerkontenZuString, EmailKonto.toString) without its errors:
empty fields where Java writes "null". Java's readers drop empty fields
(StringTokenizer) and a trailing empty one (String.split); the files are
written as a correct reader, this package's, reads them. Java escapes only
a message's text; a character another field cannot carry (a separator, a
line break) becomes a space there, with a warning, and the beans keep it.
Checked by Java: every corpus file, imported, exported and written, loads
with the e-mail programs of the original up to named departures: a string
recipient read as an address (as the writer does), the sent folder's
messages sent (the import's decision), the account files written anew
with Java's types and without the line break older versions ended the
client's with. In the synthetic model Java reads both account files back
as the model means them, every escaped character of the text included.
17 mutation probes, each loaded by Java, were all caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
exportProject.ts, step X4a: each installed program of a PC or notebook as
the FILIUS application of its class, keyed by that class name as Java keys
them, with its settings: servers switched on as the model has them, a web
server's port and virtual hosts, the DNS server's recursion on port 53, the
echo server's port, the generic client's target (none as null, as its
constructor leaves it), the peer count, the personal firewall. E-mail
contents are the next step.
The model lost a web server's virtual hosts switch (useVHost, on for 3 web
servers in the corpus); Swift sets it off on import. The core now has
WebServerConfiguration.virtualHosts (Swift's name), the import carries
Java's value, and the Swift comparison names Swift's error.
Checked by Java: every corpus file, imported, exported and written, loads
with the programs of the original, with no new departure. In the synthetic
model the PC has every program with settings away from the defaults, the
notebook programs without settings (Java's defaults, 3 peers), where Java
makes /peer2peer itself. 12 mutation probes, each loaded by Java, were all
caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
exportProject.ts, step X3: the files of a PC or notebook from the model's
virtual file system under Java's root. A file's type follows from its kind:
text by a known extension (else "text"), an image or binary data by its
media type. Images and binary data are Base64 as FILIUS's
Base64.encodeBytes (iHarder 2.x) writes it: a line break after every 76
characters, also after the last full line. Without that, all 31 images of
the corpus came back with other content. A new oracle tool, Base64Cases,
has Java encode 19 lengths around the line length; javaBase64 matches all.
A router or home router keeps the files its constructor makes.
Checked by Java: every corpus file, imported, exported and written, loads
with the files of the original, byte for byte, up to named departures of
the model: files in its order (directories first, then by name), with the
type of their kind rather than the one the file named (Java itself names
them unevenly), and no router files beyond the constructor's (www.conf,
16 times). The synthetic model adds nested directories, text with and
without a known extension, a PNG, a JPEG, binary data with and without a
media type and 57 bytes, one Base64 line exactly. 10 mutation probes, each
loaded by Java, were all caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
exportProject.ts, step X2: a switch's SSID and learning time, a host's
SSID, manual routes, RIP, DHCP client and server with static assignments, a
home router's port forwarding, the firewall of a router or home router (the
core's names as Java's numbers), and modems. What the model has no setting
for is what Java's constructors make. A remote link pair becomes modems on
the port its identifier names (12345 when it names none); the first of a
pair in node order serves, the others connect to localhost.
Checked by Java: every corpus file, imported, exported and written, loads
with the system software of the original up to two new named departures: a
switch SSID Java rolls is the one the model derives from the node id, and
the model keeps a modem pair, not which one serves (modem_beispiel.fls has
the client first; both ways work within one project). The synthetic model
now has settings away from every default, a home router with port
forwarding and two modem pairs. 21 mutation probes, each loaded by Java,
were all caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
exportProject.ts is the counterpart of importProject.ts: from the editor
model to the Java view topologyWriter.ts writes. Step X1, the topology:
each node becomes the Java device of its kind with name, position and label
flags; its ports become interfaces or ports; each link a cable drawn
between its nodes; each documentation item a GUIDocuItem. A port's MAC is
the one the simulation uses (the file's, or one derived from the port id);
a router's gateway and DNS server go to every interface, as Java's own
setters do (Java reads the first interface's only). A host with a wireless
card joins a switch by a wireless cable, the only wireless connection Java
makes. The import now enables a host's wireless by Java's card flag, also
without an SSID, where Swift enables it for every SSID.
Checked by Java: every corpus file, imported, exported and written, loads
in Java FILIUS to the topology of the original up to named departures of
the model (a derived MAC where Java rolls one at each load; documentation in
render order, with the model's empty text, Dialog 12 and minimum size where
Java has none or less). A synthetic model adds a label with IP and MAC,
positions between pixels, a five-port switch, a wireless notebook, a
three-interface router and a half-transparent bold text. 17 mutation
probes, each loaded by Java, were all caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A notebook that joins a switch's wireless network has its card marked
wireless (NetzwerkInterface.setWireless, Betriebssystem.wireless()), next
to the wireless cable. The reader did not read it, the writer did not write
it, and the Java oracle did not report it, so the Java-against-Java writer
test could not see the loss. Found while exporting the editor model: the
corpus has 18 wireless cards, one at each wireless cable.
The oracle now reports `wireless` per interface; FlsInterface has it; the
writer sets it. The synthetic topology has a wireless card; a mutation
probe that drops it is caught by Java.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
topologyWriter.ts writes the e-mail server's domain and bean accounts with
their mailboxes, and the e-mail client's three folders and address book.
The client's account is no bean (FILIUS 2.14 has no kontoListe property):
it lives in /konten.txt and is written with the files.
A Java error is not copied: older files keep recipients as strings, which
Java puts into its typed lists unchecked, and its e-mail window fails on
them. The writer writes the AddressEntry Java's own AddressEntry(String)
makes of each; a null recipient is left out. The Java oracle now reports
that reading next to each such string, and our addressEntry matches it on
all 9 in the corpus. The test names the departure (RECIPIENTS_AS_STRINGS)
and checks that it touches man-in-the-middle.fls alone.
Checked by Java against Java: every corpus file, read and written again,
loads with the same e-mail programs as the original, up to that departure.
The synthetic topology adds an e-mail server and client with an account, a
mailbox, three folders, an address book and recipients of every form. 15
mutation probes, each loaded by Java, were all caught. The writer now
writes everything the reader reads of Java's view.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
topologyWriter.ts puts each installed application of a PC or notebook under
its key and gives it the system software, as installApp does, with its
settings: a server's switch and port, a client's target, DNS recursion,
virtual hosts, the peer count, the personal firewall. The applications
follow the files, as in Java's own files: the web server, the peer-to-peer
application and the e-mail server make their directories only where the
files lack them, and the DNS server reads /dns/hosts when it is installed.
E-mail accounts, messages and address books are the next step.
The Java oracle now reports a personal firewall's rules in full, not their
number (the corpus has none; its result is byte-equal).
Checked by Java against Java: every corpus file, read and written again,
loads with the same applications as the original. The synthetic topology
adds every FILIUS application with settings away from its defaults, a
personal firewall with a rule, an own index.html (Java makes no second) and
a missing /peer2peer directory, which Java makes at the end. 12 mutation
probes, among them files and applications in reverse order, each loaded by
Java, were all caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
topologyWriter.ts adds each node's file tree to the root its constructor
makes, as XMLEncoder does (arbeitsVerzeichnis, files as new Datei(name,
typ, inhalt)); a root of its own is set anew.
The reader now reads the files of routers and home routers too: their
constructor makes files from templates (the web interface), and the file
adds to them (a web server's www.conf). systemSoftware.ts keeps what the
file adds, and templateFiles says whether the constructor's files stay in
front (false when the file sets a root of its own). The topology test finds
our part at the end of Java's tree and the template in front, the same for
every router of a class across the corpus.
Checked by Java against Java: every corpus file, read and written again,
loads with the same file systems as the original. The synthetic topology
adds a nested tree with a renamed root, a nameless directory, a file
without type, content with CR, CRLF and beyond-BMP characters, router files
after the template and a home router with a root of its own. 9 mutation
probes, each loaded by Java, were all caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
topologyWriter.ts now sets the system software each device's constructor
makes, through its setters: a switch's SSID and learning time, a host's
SSID, RIP, manual routes, DHCP client and server with static assignments,
static NAT, the modem, and a router's or home router's firewall. The
firewall is reached as XMLEncoder reaches it, through the web server's
`konfig` plug-in: the constructor hangs the same object there and among
the applications, where its key differs (Firewall on a router, NatGateway
on a home router). A switch's SSID the file lacked stays unwritten, so Java
rolls one as it did. Files and applications are later layers.
javaBeansWriter.ts: a call can carry statements on its result (a map's
value changed through get(key)), checked by Java's XMLDecoder in a new
case; a short() helper for Java's short values.
Checked by Java against Java: every corpus file, read and written again,
loads with the same system software (without files and applications) as
the original. The synthetic topology adds what the corpus lacks: a router
firewall switched on with rules of every protocol (one with a null
address), every flag away from its default, a modem with its own port and
target, and a PC's SSID, routes and DHCP server. 16 mutation probes, each
rewritten and loaded by Java, were all caught.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
topologyWriter.ts writes konfiguration.xml from Java's view of a project:
the version line, the node items (label position, device with name,
host label flags, interfaces or ports), the cable items and the
documentation items. What a device's constructor makes is changed through
getters, as XMLEncoder does, and more is added; fewer than it makes set the
list anew. Cable ends name the port objects the getters return (the writer
now names a getter's result, <void id="Port0" property="port"/>), and each
cable carries the nodes the editor draws it between.
Checked by Java against Java: every corpus file, read and written again,
loads in Java FILIUS (the oracle's loading method, --sha256 ties its result
to the written bytes) to the same devices, names, positions, interfaces,
cables and documentation as the original; so does a synthetic topology with
what the corpus never sets. The oracle and the reader now also report a
cable's drawing ends (JCablePanel ziel1/ziel2): in the corpus the nodes it
joins, for all 183 cables. Mutation probes rewrite the files and let Java
load each mutant.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
javaBeansWriter.ts writes what java.beans.XMLDecoder reads — the
counterpart of javabeans.ts and the base of the .fls writer. Its input says
explicitly what Java should do: construct a class with typed arguments, call
setters, getters (with statements on their result), methods and indexed
access; an object reached twice is written once with an id and referenced
after. Unlike XMLEncoder it writes what it is given, defaults included.
What XML 1.0 cannot hold (controls, lone surrogates, U+FFFE/F) and a CR,
which XML would read as a line break, are written as <char code>.
Checked by Java itself: nine files of the writer (strings with every kind
of odd unit, all primitive types including NaN, -0.0 and a long beyond
2^53, collections, maps, shared objects, Rectangle, Color, FILIUS beans with
getter chains, arrays, indexed access, a static constant) are read by
XMLDecoder in the Java oracle (BeansDump), which describes what it got and
collects the errors the decoder would only log. Mutation probes regenerate
the files and let Java read each mutant.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The Swift oracle reads the 43 account files Java read (--email-files, with
TopologyRuntimeEmailStorage.decodeJavaServer and decodeJavaClient). The
accounts the model takes match Swift's except where Swift checks names,
hosts and addresses and fails the whole file on one it rejects, or reads
the format otherwise (the first client line instead of the last, exactly
nine fields, a required messages field). Every departure is named with its
reason; Java's reading is right in each, none is a Java error Swift fixes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The e-mail client and server of every host come into the model (core:
EmailClientConfiguration, EmailServerConfiguration, EmailMessage), from
Java's view:
- the server's accounts are those of /mailserver/konten.txt, as Java reads
them at every simulation start; without the file those of the bean list,
which Java would drop;
- the client's account is that of /konten.txt; without it the first of an
older file's kontoListe, which FILIUS 2.14 no longer reads;
- messages are numbered per program in folder order, as Swift does; the
sent folder's are sent; a date is taken only when the file gives one.
Departures from Swift: accounts and messages stay where Swift drops what
fails its checks; first and last name stay apart; a port that is no number
is the default, with a warning. On the corpus the model matches Swift's on
all 13 files; the one Swift error is named (it numbers a mailbox's messages
as their received time and writes that into the konten.txt it rewrites).
The Swift dump now carries the e-mail state, and the file system
comparison leaves out only Swift's hidden state files (and the directories
it makes for them) instead of all of /mail, /mailserver and /peer2peer.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Java stays the reference for the .fls format, but its errors are no longer
copied. The Java oracle still measures Java as it is; each deliberate
departure is named in the tests with its reason (JAVA_BUGS) and must really
depart. The corpus - what Java itself writes - still reads exactly as in
Java.
- /dns/hosts: a label may start with a digit, holds no backslash, and the
last one is not all digits (RFC 1123, RFC 3696); A, MX and NS are read in
any case (RFC 4343); a negative TTL is invalid, so 3600 (RFC 2181). Also
for the e-mail server's domain.
- /mailserver/konten.txt: empty fields count (Java's StringTokenizer skips
them and shifts the rest, so an empty password breaks the account); a
short line alone is left out, with a warning (Java throws for the whole
file); a message keeps an empty text or subject (Java's split loses it);
a ";" in an unescaped subject stays in the message; CRLF lines.
- /konten.txt: the last complete line is the account, also with an empty
address (Java takes the last line and loses the account); CRLF lines.
- Static DHCP assignments: a broken entry alone is left out (Java loses
every entry after it).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The Java view now reads what FILIUS 2.14 keeps for its e-mail programs:
the server's domain, switch and bean accounts with their mailboxes, and the
accounts of /mailserver/konten.txt, which Java reads at every simulation
start and which so win over the bean list; the client's three folders and
address book, and its account from /konten.txt (no bean holds it).
javaEmail.ts ports the bean setters and both file readers
(stringZuListeBenutzerkonten, EmailKonto(String)); javaString.ts Java's
trim, split and StringTokenizer; javaFileSystem.ts the file lookup, now
shared with the DNS server, with the four non-ASCII units Java's
equalsIgnoreCase also takes for an ASCII letter (measured over the BMP).
Checked against Java on the corpus (the oracle dumps the e-mail state) and
on 43 account files read by EmailServer.kontenLaden and
EmailAnwendung.laden (EmailFileCases). Found: older files keep recipients
as strings in the typed lists; a getter copy of the sender is filled and
set back; one short server line makes Java throw for the whole file; a
message with an empty text is lost on reading; a broken last client line
empties the account.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Java's DNS server reads its records from the text file /dns/hosts on every
query. dnsHosts.ts ports that reading (DNSServer.leseRecordListe,
ResourceRecord, Dateisystem.holeDatei) and keeps in the model the records
Java answers: type exactly A, MX or NS, and a name Java accepted. The file
stays in the file system as it is.
Both oracles read 76 hand-picked hosts files (dns-hosts.cases.txt): the
Java oracle through DNSServer.holeResourceRecords (DnsHostsCases), the
Swift oracle through TopologyDNSHostsFileCodec (--dns-hosts). The Java view
matches Java on every case; the model matches Swift except for 33 named
cases in twelve reasons: Java splits at single spaces and LF only, has no
comments or "IN", turns a TTL that is no int into 3600, keeps duplicates,
and empties names with a leading digit. The Swift dump now carries each
server's records; the corpus comparison covers them on 16 files.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
importProject carries a host's installed programs into the model with
the settings Swift keeps (web server port, Gnutella peer limit, DNS
recursion and forwarder, personal firewall) and two it does not (the echo
server's port, the generic client's target), plus Java's "switched on".
The Swift oracle writes programs and settings; they match on the 13 files
Swift reads completely.
Intended differences, named in the test:
- every FILIUS program is kept; Swift drops command prompt, text
editor, file explorer, image viewer, echo server and generic client;
- a web server's files stay in /webserver, its document root; Swift
moves them to /www. No program state goes into a host's files.
- multicast_server.fls: maxTeilnehmerZahl refers to a shared Integer
(a button's icon text gap, 4); Swift falls back to 3.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
PeerToPeerAnwendung's constructor rolls its peer limit (3, 4 or 5). The
getter defaults measured it once each run — 4, then 5 — and c48d15d
committed 5 as a default. GetterDefaults now constructs every class 64
times and writes a value that varies as {"random": [...]}; the decoder
answers a reference to it as unresolved, and the topology layer reports
a peer limit the file does not set as null.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
readSystemSoftware lists a PC's or notebook's installed applications with
the settings the model needs: server applications switched on and their
port, the generic client's target, DNS recursion, web server virtual
hosts, Gnutella's peer limit and the personal firewall (on by default,
unlike a router's). Defaults the file leaves out are measured in Java.
All 23 corpus files match Java.
Two examples install an application of their own (Accumulator,
Broadcast), compiled into projekt/anwendungen; it is left out with a
warning, as this reader cannot run Java classes. It must survive a save.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
importProject turns a PC's or notebook's Java file tree into Swift's
virtual file system, as Swift's registerVirtualFileSystem does: images
by type or extension from Base64, text by type, unknown types as binary
data or text, unsafe names and collisions left out with a warning, and
the whole import rejected over the project quota. NSString.pathExtension
is measured in the oracle, which now writes every imported file system.
On the 13 files Swift reads completely every file matches, images
compared by hash. Intended: a host whose file has no file system stays
empty, as in Java, where Swift adds the demo files of a new device.
Pending the applications step, the comparison reads our /webserver as
Swift's /www and leaves out the e-mail and Gnutella state files.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The core gains Swift's virtual file system (paths, case-insensitive
lookup that keeps the spelling, quotas, atomic changes, the defaults of
a new device) and the string behaviour the ported code relies on without
saying so. Both are measured in the Swift oracle: 11 file system cases
(the three Swift tests among them) step by step, and trimming, lower
case, equality, ordering and Base64 case by case — whitespace and control
characters over all of Unicode.
Found by measuring:
- Swift trims NEL and ZERO WIDTH SPACE, not the BOM; JS trim() does the
opposite. The .fls import now trims with swiftTrim.
- lowercased() maps scalar by scalar: a word-final sigma stays σ.
- Foundation decodes "====" and "Y===" as empty data.
- Foundation on Linux counts 24,735 plane-14 scalars as control
characters, a table artefact; the port keeps Unicode Cc+Cf.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
readSystemSoftware reads a host's file system: Java's tree of
DefaultMutableTreeNodes, directories named by their user object, files
as Datei (by setters or the three-argument constructor), rooted at
arbeitsVerzeichnis (up to 1.10) or root. All 23 corpus files match Java,
every file compared by name, type and the SHA-256 of its content.
Routers and home routers have files in Java too, but their constructor
makes them from templates (the start page of the firewall's web
interface); they are not in the file and not claimed here.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
importProject carries the DHCP client switch and server settings (hosts
and home routers), a home router's port forwarding rows and a modem's
pairing into the model; the core gains the types, ported from Swift.
The Swift oracle writes them too, and they match on the 13 files Swift
reads completely. Configuration text is now trimmed throughout, as Swift
trims every value it reads (routes and firewall rules included).
Outside those files Swift switches every DHCP server of the Internet
examples off: their "aktiv" refers to a shared Boolean it cannot follow.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
readSystemSoftware adds the DHCP client switch and server settings of
every IP stack, a home router's static NAT table (port forwarding) and a
modem's mode, port and address. All 23 corpus files match Java.
Static DHCP assignments follow setStaticAssignedAddresses: each entry is
split at Java whitespace (the same 25 characters as Character.isWhitespace,
measured over the BMP), and an entry without MAC and IP ends the list, as
Java throws there.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
importProject carries the system software settings into the model the
core runs: switch SSID and learning time (an SSID Java would roll comes
from the node id, as in Swift), a host's wireless network, manual routes,
RIP and the firewall of every router and home router. The Swift oracle
writes these settings too; they match on the 13 files Swift reads
completely.
A router's default gateway — its first interface's, which Java routes to
(Weiterleitungstabelle.holeTabelle) — is imported into the router's
device configuration. Swift's model drops it; in internetwork_komplett
the way back from 141.99.5.10 depends on it. A test pings across three
routers in the core, with and without it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
readSystemSoftware adds to each node what the topology needs from its
system software: a switch's SSID and learning time, a host's wireless
SSID, a router's RIP switch, the manual routes and the firewall of a
router or home router. All 23 corpus files match Java.
- A router's firewall is one object its constructor hangs in two places;
the corpus writes it under the web server's "konfig" plug-in. The
places are merged.
- The decoder keeps entries a constructor made in a map, reached by
get(key) (installed applications), like the elements of a list.
- The Java oracle loads every file twice and writes values that differ
as null: what FILIUS rolls (MACs, a switch's SSID) without text search.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
importProject builds what the simulation core runs from Java's view of a
file: nodes with kinds and ports, links between port ids, device and
interface IP configurations, documentation items with Swift's derived
ids. The core gains the model types it needs (node, link, graph,
documentation item), ported from filius-ipad.
Checked against the Swift import, which the Swift oracle now writes in
full (corpus.swift.json, "state"), on the 13 files Swift reads
completely. Two differences are intended and named in the test:
- a device the file leaves unnamed has Java's default name
("Switch / WLAN"), not Swift's;
- multicast_server.fls: useIPAsName refers to a shared Boolean that
Swift cannot follow; Java shows the IP address, and so do we.
All 23 files import with as many devices and cables as Java has, and an
imported example pings across its router in the core.
Port counts of switches and routers follow Java's lists instead of
Swift's guess from references.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
readTopology maps the decoded JavaBeans graph to devices, cables and
documentation items the way SzenarioVerwaltung.netzwerkLaden does: the
kind from the Java class (the palette label is not file data), devices
without bounds at 0,0, interface values through Java's setters. All 23
corpus files match Java field by field.
Found on the way:
- listItems mixed a constructor's elements with added ones; a router's
second interface was lost. Lists now replay add/set in order after the
elements the constructor made.
- The first 0 of a file may be SwingConstants.CENTER; measured like the
getter defaults.
- IP.ipCheck (javaIp.ts) measured on 46 cases: empty tokens skipped,
signs, leading zeros and BMP digits of any script accepted.
- The Java oracle took lower-case MACs in a file for random ones (8
interfaces); it now compares case-insensitively and reports the host
name flags.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A generic interpreter for the .fls configuration: objects, properties,
calls, indexes, static fields and factories, Field.set, and id/idref
anywhere — including getter results, answered from the measured Java
defaults or reported as unresolved. All 23 corpus files: class and
position of every node as Java reads them, also where Swift loses nodes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
XMLEncoder shares equal Integer objects, so a node's x coordinate can be
written as a reference to the result of JSidebarButton.getIconTextGap().
GetterDefaults constructs the classes like XMLDecoder does and records
what the getters return (packages/fls/src/javaDefaults.json).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
41 path pairs replayed in the Swift core decide which spellings name
the same file. Two corrections: full Unicode case folding (ß, ligatures,
final sigma, micro, long s) instead of toLowerCase, and width folding of
the whole Halfwidth and Fullwidth Forms block. Swift treats composed and
decomposed spellings as one dictionary key; here the writer refuses both.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Reader and writer ported from filius-ipad with the same path policy,
limits and error codes; DEFLATE via fflate, cut at the declared size
like Apple Compression. 54 tests: the Swift archive rejection tests,
DEFLATE edge cases, and all 23 corpus files cross-checked against
fflate.unzipSync and round-tripped.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
tools/java-oracle builds FILIUS at 6640354 and loads every .fls through
its own private loader under Xvfb. All 23 files load with every node
positioned and every cable resolved (Swift: 20 of 25 nodes in the
Internet example). MAC addresses FILIUS generates are left out, so two
runs are byte-identical.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The oracle gains a corpus mode; run.sh now stops on a failed build
instead of running a stale binary. Swift reads all 23 files but drops
Java gateways labelled 'Heimrouter' and nodes whose bounds are written
through XMLEncoder back-references.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
tools/swift-oracle builds the unchanged Swift runtime (pinned 3b5ded0)
in Docker and replays JSON scenarios; the TypeScript engine must produce
the same trace events, results and end state. Five scenarios, 884
events, identical.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
First slice of the TypeScript port of the filius-ipad runtime engine,
with 29 tests ported from the Swift suite. Parts not ported yet throw
NotPortedError instead of guessing.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
